THOMPSON, SECRETARY OF HEALTH AND HUMAN SERVICES, et al. v. WESTERN STATES MEDICAL CENTER et al.

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
357 OCTOBER TERM, 2001
Syllabus
THOMPSON, SECRETARY OF HEALTH AND HUMAN
SERVICES, et al. v. WESTERN STATES
MEDICAL CENTER et al.
certiorari to the united states court of appeals for
the ninth circuit
No. 01–344. Argued February 26, 2002—Decided April 29, 2002
Drug compounding is a process by which a pharmacist or doctor combines,
mixes, or alters ingredients to create a medication tailored to an individ-
ual patient’s needs. The Food and Drug Administration Modernization
Act of 1997 (FDAMA) exempts “compounded drugs” from the Food and
Drug Administration’s (FDA) standard drug approval requirements
under the Federal Food, Drug, and Cosmetic Act (FDCA), so long as
the providers of the compounded drugs abide by several restrictions,
including that the prescription be “unsolicited,” 21 U. S. C. § 353a(a), and
that the providers “not advertise or promote the compounding of any
particular drug, class of drug, or type of drug,” § 353a(c). Respondents,
a group of licensed pharmacies that specialize in compounding drugs,
sought to enjoin enforcement of the advertising and solicitation pro-
visions, arguing that they violate the First Amendment’s free speech
guarantee. The District Court agreed and granted respondents sum-
mary judgment, holding that the provisions constitute unconstitutional
restrictions on commercial speech under Central Hudson Gas & Elec.
Corp. v. Public Serv. Comm’n of N. Y., 447 U. S. 557, 566. Affirming in
relevant part, the Ninth Circuit held that the restrictions in question
fail Central Hudson’s test because the Government had not demon-
strated that the restrictions would directly advance its interests or that
alternatives less restrictive of speech were unavailable.
Held: The FDAMA’s prohibitions on soliciting prescriptions for, and ad-
vertising, compounded drugs amount to unconstitutional restrictions on
commercial speech. Pp. 366–377.
(a) For a commercial speech regulation to be constitutionally permis-
sible under the Central Hudson test, the speech in question must con-
cern lawful activity and not be misleading, the asserted governmental
interest to be served by the regulation must be substantial, and the
regulation must “directly advanc[e]” the governmental interest and “not
[be] more extensive than is necessary to serve that interest,” 447 U. S.,
at 566. Pp. 366–368.
(b) The Government asserts that three substantial interests underlie
the FDAMA: (1) preserving the effectiveness and integrity of the

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
358 THOMPSON v. WESTERN STATES MEDICAL CENTER
Syllabus
FDCA’s new drug approval process and the protection of the public
health it provides; (2) preserving the availability of compounded drugs
for patients who, for particularized medical reasons, cannot use com-
mercially available products approved by the FDA; and (3) achieving
the proper balance between those two competing interests. Preserving
the new drug approval process is clearly an important governmental
interest, as is permitting the continuation of the practice of com-
pounding so that patients with particular needs may obtain medica-
tions suited to those needs. Because pharmacists do not make enough
money from small-scale compounding to make safety and efficacy testing
of their compounded drugs economically feasible, however, it would not
make sense to require compounded drugs created to meet the unique
needs of individual patients to undergo the entire new drug approval
process. The Government therefore needs to be able to draw a line
between small-scale compounding and large-scale drug manufacturing.
The Government argues that the FDAMA’s speech-related provisions
provide just such a line: As long as pharmacists do not advertise particu-
lar compounded drugs, they may sell compounded drugs without first
undergoing safety and efficacy testing and obtaining FDA approval.
However, even assuming that the FDAMA’s prohibition on advertising
compounded drugs “directly advance[s]” the Government’s asserted
interests, the Government has failed to demonstrate that the speech
restrictions are “not more extensive than is necessary to serve [those]
interest[s].” Central Hudson, supra, at 566. If the Government can
achieve its interests in a manner that does not restrict commercial
speech, or that restricts less speech, the Government must do so.
E. g., Rubin v. Coors Brewing Co., 514 U. S. 476, 490–491. Several
non-speech-related means of drawing a line between compounding and
large-scale manufacturing might be possible here. For example, the
Government could ban the use of commercial scale manufacturing or
testing equipment in compounding drug products, prohibit pharmacists
from compounding more drugs in anticipation of receiving prescrip-
tions than in response to prescriptions already received, or prohibit
them from offering compounded drugs at wholesale to other state
licensed persons or commercial entities for resale. The Government
has not offered any reason why such possibilities, alone or in combi-
nation, would be insufficient to prevent compounding from occurring
on such a scale as to undermine the new drug approval process.
Pp. 368–373.
(c) Even if the Government had argued (as does the dissent) that the
FDAMA’s speech-related restrictions were motivated by a fear that ad-
vertising compounded drugs would put people who do not need such
drugs at risk by causing them to convince their doctors to prescribe the
drugs anyway, that fear would fail to justify the restrictions. This

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
359 Cite as: 535 U. S. 357 (2002)
Syllabus
concern rests on the questionable assumption that doctors would pre-
scribe unnecessary medications and amounts to a fear that people would
make bad decisions if given truthful information, a notion that the Court
rejected as a justification for an advertising ban in, e. g., Virginia Bd.
of Pharmacy v. Virginia Citizens Consumer Council, Inc., 425 U. S.
748, 770. Pp. 373–376.
(d) If the Government’s failure to justify its decision to regulate
speech were not enough to convince the Court that the FDAMA’s ad-
vertising provisions were unconstitutional, the amount of beneficial
speech prohibited by the FDAMA would be. Forbidding the adver-
tisement of compounded drugs would prevent pharmacists with no
interest in mass-producing medications, but who serve clienteles with
special medical needs, from telling the doctors treating those clients
about the alternative drugs available through compounding. For exam-
ple, a pharmacist serving a children’s hospital where many patients are
unable to swallow pills would be prevented from telling the children’s
doctors about a new development in compounding that allowed a drug
that was previously available only in pill form to be administered an-
other way. The fact that the FDAMA would prohibit such seemingly
useful speech even though doing so does not appear to directly further
any asserted governmental objective confirms that the prohibition is
unconstitutional. Pp. 376–377.
238 F. 3d 1090, affirmed.
O’Connor, J., delivered the opinion of the Court, in which Scalia, Ken-
nedy, Souter, and Thomas, JJ., joined. Thomas, J., filed a concurring
opinion, post, p. 377. Breyer, J., filed a dissenting opinion, in which
Rehnquist, C. J., and Stevens and Ginsburg, JJ., joined, post, p. 378.
Deputy Solicitor General Kneedler argued the cause for
petitioners. With him on the briefs were Solicitor General
Olson, Assistant Attorney General McCallum, Matthew D.
Roberts, Douglas N. Letter, Alex M. Azar II, Daniel E. Troy,
and Patricia J. Kaeding.
Howard M. Hoffmann argued the cause and filed a brief
for respondents.*
*Briefs of amici curiae urging affirmance were filed for the Interna-
tional Academy of Compounding Pharmacists by Alan E. Untereiner and
Arnon D. Siegel; for the National Community Pharmacists Association by
Kenneth S. Geller and John M. Rector; and for Julian M. Whitaker, M.D.,
et al. by Jonathan W. Emord and Claudia A. Lewis-Eng.
Michael H. McConihe filed a brief for the American Pharmaceutical
Association as amicus curiae.

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
360 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
Justice O’Connor delivered the opinion of the Court.
Section 127(a) of the Food and Drug Administration Mod-
ernization Act of 1997 (FDAMA or Act), 111 Stat. 2328,
21 U. S. C. § 353a, exempts “compounded drugs” from the
Food and Drug Administration’s standard drug approval re-
quirements as long as the providers of those drugs abide
by several restrictions, including that they refrain from
advertising or promoting particular compounded drugs.
Respondents, a group of licensed pharmacies that special-
ize in compounding drugs, sought to enjoin enforcement
of the subsections of the Act dealing with advertising and
solicitation, arguing that those provisions violate the First
Amendment’s free speech guarantee. The District Court
agreed with respondents and granted their motion for sum-
mary judgment, holding that the provisions do not meet the
test for acceptable government regulation of commercial
speech set forth in Central Hudson Gas & Elec. Corp. v.
Public Serv. Comm’n of N. Y., 447 U. S. 557, 566 (1980). The
court invalidated the relevant provisions, severing them
from the rest of § 127(a).
The Court of Appeals for the Ninth Circuit affirmed in
part and reversed in part, agreeing that the provisions re-
garding advertisement and promotion are unconstitutional
but finding them not to be severable from the rest of § 127(a).
Petitioners challenged only the Court of Appeals’ constitu-
tional holding in their petition for certiorari, and respond-
ents did not file a cross-petition. We therefore address only
the constitutional question, having no occasion to review the
Court of Appeals’ severability determination. We conclude,
as did the courts below, that § 127(a)’s provisions regard-
ing advertisement and promotion amount to unconstitutional
restrictions on commercial speech, and we therefore affirm.
I
Drug compounding is a process by which a pharmacist
or doctor combines, mixes, or alters ingredients to create

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
361 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
a medication tailored to the needs of an individual patient.
Compounding is typically used to prepare medications that
are not commercially available, such as medication for a
patient who is allergic to an ingredient in a mass-produced
product. It is a traditional component of the practice of
pharmacy, see J. Thompson, A Practical Guide to Contem-
porary Pharmacy Practice 11.3 (1998), and is taught as part
of the standard curriculum at most pharmacy schools, see
American Council on Pharmaceutical Education, Accredita-
tion Standards and Guidelines for the Professional Program
in Pharmacy Leading to the Doctor of Pharmacy Degree,
Standard 10(a) (adopted June 14, 1997). Many States spe-
cifically regulate compounding practices as part of their
regulation of pharmacies. See, e. g., Cal. Code Regs., tit. 16,
§§ 1716.2, 1751 (2002); Ind. Admin. Code, tit. 856, §§ 1–30–8,
1–30–18, 1–28–8 (2001); N. H. Code Admin. Rules Ann.
Pharmacy, pts. PH 404, PH 702.01 (2002); 22 Tex. Admin.
Code § 291.36 (2002). Some require all licensed pharma-
cies to offer compounding services. See, e. g., 49 Pa. Code
§ 27.18(p)(2) (2002); W. Va. Code St. Rules, tit. 15, § 19.4
(2002). Pharmacists may provide compounded drugs to pa-
tients only upon receipt of a valid prescription from a doctor
or other medical practitioner licensed to prescribe medica-
tion. See, e. g., Okla. Admin. Code §§ 535:15–10–3, 535:15–
10–9(d) (2001); Colo. State Board of Pharmacy Rule 3.02.10
(2001).
The Federal Food, Drug, and Cosmetic Act of 1938
(FDCA), 21 U. S. C. §§ 301–397, regulates drug manufac-
turing, marketing, and distribution. Section 505(a) of the
FDCA, 52 Stat. 1052, as amended, 76 Stat. 784, provides that
“[n]o person shall introduce or deliver for introduction into
interstate commerce any new drug, unless an approval of an
application filed [with the Food and Drug Administration] is
effective with respect to such drug.” 21 U. S. C. § 355(a).
“[N]ew drug” is defined by § 201(p)(1) of the FDCA, 52
Stat. 1041, as amended, 76 Stat. 781, as “[a]ny drug . . . not

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
362 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
generally recognized, among experts qualified by scientific
training and experience to evaluate the safety and effective-
ness of drugs, as safe and effective for use under the condi-
tions prescribed, recommended, or suggested in the labeling
thereof.” 21 U. S. C. § 321(p). The FDCA invests the Food
and Drug Administration (FDA) with the power to enforce
its requirements. § 371(a).
For approximately the first 50 years after the enactment
of the FDCA, the FDA generally left regulation of com-
pounding to the States. Pharmacists continued to provide
patients with compounded drugs without applying for FDA
approval of those drugs. The FDA eventually became con-
cerned, however, that some pharmacists were manufacturing
and selling drugs under the guise of compounding, thereby
avoiding the FDCA’s new drug requirements. In 1992,
in response to this concern, the FDA issued a Compliance
Policy Guide, which announced that the “FDA may, in the
exercise of its enforcement discretion, initiate federal en-
forcement actions . . . when the scope and nature of a phar-
macy’s activities raises the kinds of concerns normally asso-
ciated with a manufacturer and . . . results in significant
violations of the new drug, adulteration, or misbranding
provisions of the Act.” Compliance Policy Guide 7132.16
(hereinafter Guide), App. to Pet. for Cert. 76a. The Guide
explained that the “FDA recognizes that pharmacists tra-
ditionally have extemporaneously compounded and ma-
nipulated reasonable quantities of drugs upon receipt of a
valid prescription for an individually identified patient from
a licensed practitioner,” and that such activity was not the
subject of the Guide. Id., at 71a. The Guide said, however,
“that while retail pharmacies . . . are exempted from certain
requirements of the [FDCA], they are not the subject of any
general exemption from the new drug, adulteration, or mis-
branding provisions” of the FDCA. Id., at 72a. It stated
that the “FDA believes that an increasing number of es-
tablishments with retail pharmacy licenses are engaged in

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
363 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
manufacturing, distributing, and promoting unapproved new
drugs for human use in a manner that is clearly outside the
bounds of traditional pharmacy practice and that constitute
violations of the [FDCA].” Ibid. The Guide expressed
concern that drug products “manufactured and distributed
in commercial amounts without [the] FDA’s prior approval”
could harm the public health. Id., at 73a.
In light of these considerations, the Guide announced that
it was FDA policy to permit pharmacists to compound drugs
after receipt of a valid prescription for an individual patient
or to compound drugs in “very limited quantities” before re-
ceipt of a valid prescription if they could document a history
of receiving valid prescriptions “generated solely within an
established professional practitioner-patient-pharmacy re-
lationship” and if they maintained the prescription on file
as required by state law. Id., at 73a–75a. Compounding
in such circumstances was permitted as long as the phar-
macy’s activities did not raise “the kinds of concerns nor-
mally associated with a manufacturer.” Id., at 76a. The
Guide listed nine examples of activities that the FDA be-
lieved raised such concerns and that would therefore be con-
sidered by the agency in determining whether to bring an
enforcement action. These activities included: “[s]oliciting
business (e. g., promoting, advertising, or using salespersons)
to compound specific drug products, product classes, or
therapeutic classes of drug products”; “[c]ompounding, regu-
larly, or in inordinate amounts, drug products that are com-
mercially available . . . and that are essentially generic copies
of commercially available, FDA–approved drug products”;
using commercial scale manufacturing or testing equipment
to compound drugs; offering compounded drugs at wholesale;
and “[d]istributing inordinate amounts of compounded prod-
ucts out of state.” Id., at 76a–77a. The Guide further
warned that pharmacies could not dispense drugs to third
parties for resale to individual patients without losing their
status as retail entities. Id., at 75a.

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
364 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
Congress turned portions of this policy into law when it
enacted the FDAMA in 1997. The FDAMA, which amends
the FDCA, exempts compounded drugs from the FDCA’s
“new drug” requirements and other requirements provided
the drugs satisfy a number of restrictions. First, they must
be compounded by a licensed pharmacist or physician in re-
sponse to a valid prescription for an identified individual
patient, or, if prepared before the receipt of such a pre-
scription, they must be made only in “limited quantities”
and in response to a history of the licensed pharmacist’s or
physician’s receipt of valid prescription orders for that drug
product within an established relationship between the phar-
macist, the patient, and the prescriber. 21 U. S. C. § 353a(a).
Second, the compounded drug must be made from ap-
proved ingredients that meet certain manufacturing and
safety standards, §§ 353a(b)(1)(A)–(B), and the compounded
drug may not appear on an FDA list of drug products that
have been withdrawn or removed from the market because
they were found to be unsafe or ineffective, § 353a(b)(1)(C).
Third, the pharmacist or physician compounding the drug
may not “compound regularly or in inordinate amounts
(as defined by the Secretary) any drug products that are
essentially copies of a commercially available drug product.”
§ 353a(b)(1)(D). Fourth, the drug product must not be iden-
tified by the FDA as a drug product that presents demon-
strable difficulties for compounding in terms of safety or
effectiveness. § 353a(b)(3)(A). Fifth, in States that have
not entered into a “memorandum of understanding” with the
FDA addressing the distribution of “inordinate amounts” of
compounded drugs in interstate commerce, the pharmacy,
pharmacist, or physician compounding the drug may not
distribute compounded drugs out of state in quantities ex-
ceeding five percent of that entity’s total prescription orders.
§ 353a(b)(3)(B). Finally, and most relevant for this litiga-
tion, the prescription must be “unsolicited,” § 353a(a), and
the pharmacy, licensed pharmacist, or licensed physician

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
365 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
compounding the drug may “not advertise or promote the
compounding of any particular drug, class of drug, or type
of drug,” § 353a(c). The pharmacy, licensed pharmacist, or
licensed physician may, however, “advertise and promote
the compounding service.” Ibid.
Respondents are a group of licensed pharmacies that spe-
cialize in drug compounding. They have prepared promo-
tional materials that they distribute by mail and at medical
conferences to inform patients and physicians of the use and
effectiveness of specific compounded drugs. Fearing that
they would be prosecuted under the FDAMA if they con-
tinued to distribute those materials, respondents filed a
complaint in the United States District Court for the Dis-
trict of Nevada, arguing that the Act’s requirement that
they refrain from advertising and promoting their products
if they wish to continue compounding violates the Free
Speech Clause of the First Amendment. Specifically, they
challenged the requirement that prescriptions for com-
pounded drugs be “unsolicited,” 21 U. S. C. § 353a(a), and
the requirement that pharmacists “not advertise or pro-
mote the compounding of any particular drug, class of drug,
or type of drug,” § 353a(c). The District Court granted sum-
mary judgment to respondents, finding that the FDAMA’s
speech-related provisions constitute unconstitutional restric-
tions on commercial speech under Central Hudson, 447 U. S.,
at 566, and that their enforcement should therefore be en-
joined. Western States Medical Center v. Shalala, 69 F.
Supp. 2d 1288 (Nev. 1999). The District Court, however,
found those provisions to be severable from the rest of
§ 127(a) of the FDAMA, 21 U. S. C. § 353a, and so left the
Act’s other compounding requirements intact.
The Government appealed both the holding that the
speech-related provisions were unconstitutional and the
holding that those provisions were severable from the rest
of § 127(a). The Court of Appeals for the Ninth Circuit af-
firmed in part and reversed in part. Western States Med-

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
366 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
ical Center v. Shalala, 238 F. 3d 1090 (2001). The Court
of Appeals agreed that the FDAMA’s advertisement and
solicitation restrictions fail Central Hudson’s test for per-
missible regulation of commercial speech, finding that the
Government had not demonstrated that the speech restric-
tions would directly advance its interests or that alterna-
tives less restrictive of speech were unavailable. The Court
of Appeals disagreed, however, that the speech-related re-
strictions were severable from the rest of § 127(a), 21 U. S. C.
§ 353a, explaining that the FDAMA’s legislative history dem-
onstrated that Congress intended to exempt compounding
from the FDCA’s requirements only in return for a prohibi-
tion on promotion of specific compounded drugs. Accord-
ingly, the Court of Appeals invalidated § 127(a) in its entirety.
We granted certiorari, 534 U. S. 992 (2001), to consider
whether the FDAMA’s prohibitions on soliciting prescrip-
tions for, and advertising, compounded drugs violate the
First Amendment. Because neither party petitioned for
certiorari on the severability issue, we have no occasion to
review that portion of the Court of Appeals’ decision. Like-
wise, the provisions of the FDAMA outside § 127(a), which
are unrelated to drug compounding, are not an issue here
and so remain unaffected.
II
The parties agree that the advertising and soliciting pro-
hibited by the FDAMA constitute commercial speech. In
Virginia Bd. of Pharmacy v. Virginia Citizens Consumer
Council, Inc., 425 U. S. 748 (1976), the first case in which we
explicitly held that commercial speech receives First Amend-
ment protection, we explained the reasons for this protec-
tion: “It is a matter of public interest that [economic] deci-
sions, in the aggregate, be intelligent and well-informed. To
this end, the free flow of commercial information is indispen-
sable.” Id., at 765. Indeed, we recognized that a “particu-
lar consumer’s interest in the free flow of commercial
information . . . may be as keen, if not keener by far, than

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
367 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
his interest in the day’s most urgent political debate.” Id.,
at 763. We have further emphasized:
“The commercial marketplace, like other spheres of our
social and cultural life, provides a forum where ideas
and information flourish. Some of the ideas and infor-
mation are vital, some of slight worth. But the general
rule is that the speaker and the audience, not the gov-
ernment, assess the value of the information presented.
Thus, even a communication that does no more than pro-
pose a commercial transaction is entitled to the coverage
of the First Amendment.” Edenfield v. Fane, 507 U. S.
761, 767 (1993).
Although commercial speech is protected by the First
Amendment, not all regulation of such speech is unconsti-
tutional. See Virginia Bd. of Pharmacy, supra, at 770. In
Central Hudson, supra, we articulated a test for deter-
mining whether a particular commercial speech regulation
is constitutionally permissible. Under that test we ask as a
threshold matter whether the commercial speech concerns
unlawful activity or is misleading. If so, then the speech
is not protected by the First Amendment. If the speech
concerns lawful activity and is not misleading, however,
we next ask “whether the asserted governmental interest
is substantial.” Id., at 566. If it is, then we “determine
whether the regulation directly advances the governmental
interest asserted,” and, finally, “whether it is not more
extensive than is necessary to serve that interest.” Ibid.
Each of these latter three inquiries must be answered in the
affirmative for the regulation to be found constitutional.
Neither party has challenged the appropriateness of apply-
ing the Central Hudson framework to the speech-related
provisions at issue here. Although several Members of the
Court have expressed doubts about the Central Hudson
analysis and whether it should apply in particular cases,
see, e. g., Greater New Orleans Broadcasting Assn., Inc. v.

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
368 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
United States, 527 U. S. 173, 197 (1999) (Thomas, J., concur-
ring in judgment); 44 Liquormart, Inc. v. Rhode Island, 517
U. S. 484, 501, 510–514 (1996) (opinion of Stevens, J., joined
by Kennedy and Ginsburg, JJ.); id., at 517 (Scalia, J.,
concurring in part and concurring in judgment); id., at 518
(Thomas, J., concurring in part and concurring in judgment),
there is no need in this case to break new ground. “ ‘Central
Hudson, as applied in our more recent commercial speech
cases, provides an adequate basis for decision.’ ” Lorillard
Tobacco Co. v. Reilly, 533 U. S. 525, 554–555 (2001) (quoting
Greater New Orleans, supra, at 184).
III
The Government does not attempt to defend the FDAMA’s
speech-related provisions under the first prong of the Cen-
tral Hudson test; i. e., it does not argue that the prohibited
advertisements would be about unlawful activity or would
be misleading. Instead, the Government argues that the
FDAMA satisfies the remaining three prongs of the Central
Hudson test.
The Government asserts that three substantial interests
underlie the FDAMA. The first is an interest in “pre-
serv[ing] the effectiveness and integrity of the FDCA’s new
drug approval process and the protection of the public health
that it provides.” Brief for Petitioners 19. The second is
an interest in “preserv[ing] the availability of compounded
drugs for those individual patients who, for particularized
medical reasons, cannot use commercially available prod-
ucts that have been approved by the FDA.” Id., at 19–20.
Finally, the Government argues that “[a]chieving the proper
balance between those two independently compelling but
competing interests is itself a substantial governmental in-
terest.” Id., at 20.
Explaining these interests, the Government argues that
the FDCA’s new drug approval requirements are critical
to the public health and safety. It claims that the FDA’s

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
369 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
experience with drug regulation demonstrates that proof of
the safety and effectiveness of a new drug needs to be es-
tablished by rigorous, scientifically valid clinical studies be-
cause impressions of individual doctors, who cannot them-
selves compile sufficient safety data, cannot be relied upon.
The Government also argues that a premarket approval
process, under which manufacturers are required to put their
proposed drugs through tests of safety and effectiveness in
order to obtain FDA approval to market the drugs, is the
best way to guarantee drug safety and effectiveness.
While it praises the FDCA’s new drug approval process,
the Government also acknowledges that “because obtain-
ing FDA approval for a new drug is a costly process, re-
quiring FDA approval of all drug products compounded by
pharmacies for the particular needs of an individual patient
would, as a practical matter, eliminate the practice of com-
pounding, and thereby eliminate availability of compounded
drugs for those patients who have no alternative treatment.”
Id., at 26. The Government argues that eliminating the
practice of compounding drugs for individual patients would
be undesirable because compounding is sometimes critical
to the care of patients with drug allergies, patients who can-
not tolerate particular drug delivery systems, and patients
requiring special drug dosages.
Preserving the effectiveness and integrity of the FDCA’s
new drug approval process is clearly an important gov-
ernmental interest, and the Government has every reason
to want as many drugs as possible to be subject to that
approval process. The Government also has an important
interest, however, in permitting the continuation of the
practice of compounding so that patients with particular
needs may obtain medications suited to those needs. And
it would not make sense to require compounded drugs cre-
ated to meet the unique needs of individual patients to
undergo the testing required for the new drug approval
process. Pharmacists do not make enough money from

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
370 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
small-scale compounding to make safety and efficacy test-
ing of their compounded drugs economically feasible, so re-
quiring such testing would force pharmacists to stop provid-
ing compounded drugs. Given this, the Government needs
to be able to draw a line between small-scale compounding
and large-scale drug manufacturing. That line must distin-
guish compounded drugs produced on such a small scale that
they could not undergo safety and efficacy testing from drugs
produced and sold on a large enough scale that they could
undergo such testing and therefore must do so.
The Government argues that the FDAMA’s speech-related
provisions provide just such a line, i. e., that, in the terms
of Central Hudson, they “directly advanc[e] the govern-
mental interest[s] asserted.” 447 U. S., at 566. Those pro-
visions use advertising as the trigger for requiring FDA
approval—essentially, as long as pharmacists do not ad-
vertise particular compounded drugs, they may sell com-
pounded drugs without first undergoing safety and efficacy
testing and obtaining FDA approval. If they advertise
their compounded drugs, however, FDA approval is re-
quired. The Government explains that traditional (or, in
its view, desirable) compounding responds to a physician’s
prescription and an individual patient’s particular medical
situation, and that “[a]dvertising the particular products
created in the provision of [such] service (as opposed to ad-
vertising the compounding service itself) is not necessary
to . . . this type of responsive and customized service.”
Brief for Petitioners 34. The Government argues that ad-
vertising particular products is useful in a broad market
but is not useful when particular products are designed in
response to an individual’s “often unique need[s].” Ibid.
The Government contends that, because of this, advertising
is not typically associated with compounding for particular
individuals. In contrast it is typically associated, the Gov-
ernment claims, with large-scale production of a drug for a
substantial market. The Government argues that advertis-

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
371 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
ing, therefore, is “a fair proxy for actual or intended large-
scale manufacturing,” and that Congress’ decision to limit
the FDAMA’s compounding exemption to pharmacies that do
not engage in promotional activity was “rationally calcu-
lated” to avoid creating “ ‘a loophole that would allow un-
regulated drug manufacturing to occur under the guise of
pharmacy compounding.’ ” Id., at 35 (quoting 143 Cong.
Rec. S9839 (Sept. 24, 1997) (statement of Sen. Kennedy)).
The Government seems to believe that without advertising
it would not be possible to market a drug on a large enough
scale to make safety and efficacy testing economically fea-
sible. The Government thus believes that conditioning an
exemption from the FDA approval process on refraining
from advertising is an ideal way to permit compounding
and yet also guarantee that compounding is not conducted
on such a scale as to undermine the FDA approval process.
Assuming it is true that drugs cannot be marketed on a
large scale without advertising, the FDAMA’s prohibition
on advertising compounded drugs might indeed “directly
advanc[e]” the Government’s interests. Central Hudson,
447 U. S., at 566. Even assuming that it does, however, the
Government has failed to demonstrate that the speech re-
strictions are “not more extensive than is necessary to
serve [those] interest[s].” Ibid. In previous cases address-
ing this final prong of the Central Hudson test, we have
made clear that if the Government could achieve its interests
in a manner that does not restrict speech, or that restricts
less speech, the Government must do so. In Rubin v. Coors
Brewing Co., 514 U. S. 476 (1995), for example, we found a
law prohibiting beer labels from displaying alcohol content
to be unconstitutional in part because of the availability
of alternatives “such as directly limiting the alcohol con-
tent of beers, prohibiting marketing efforts emphasizing high
alcohol strength . . . , or limiting the labeling ban only to
malt liquors.” Id., at 490–491. The fact that “all of [these
alternatives] could advance the Government’s asserted inter-

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
372 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
est in a manner less intrusive to . . . First Amendment
rights” indicated that the law was “more extensive than nec-
essary.” Id., at 491. See also 44 Liquormart, Inc. v. Rhode
Island, 517 U. S., at 507 (plurality opinion) (striking down a
prohibition on advertising the price of alcoholic beverages
in part because “alternative forms of regulation that would
not involve any restriction on speech would be more likely
to achieve the State’s goal of promoting temperance”).
Several non-speech-related means of drawing a line be-
tween compounding and large-scale manufacturing might
be possible here. First, it seems that the Government could
use the very factors the FDA relied on to distinguish com-
pounding from manufacturing in its 1992 Guide. For exam-
ple, the Government could ban the use of “commercial scale
manufacturing or testing equipment for compounding drug
products.” Guide, App. to Pet. for Cert. 76a. It could pro-
hibit pharmacists from compounding more drugs in antic-
ipation of receiving prescriptions than in response to pre-
scriptions already received. See ibid. It could prohibit
pharmacists from “[o]ffering compounded drug products at
wholesale to other state licensed persons or commercial enti-
ties for resale.” Id., at 77a. Alternately, it could limit the
amount of compounded drugs, either by volume or by num-
bers of prescriptions, that a given pharmacist or pharmacy
sells out of state. See ibid. Another possibility not sug-
gested by the Guide would be capping the amount of any
particular compounded drug, either by drug volume, number
of prescriptions, gross revenue, or profit that a pharmacist
or pharmacy may make or sell in a given period of time. It
might even be sufficient to rely solely on the non-speech-
related provisions of the FDAMA, such as the requirement
that compounding only be conducted in response to a pre-
scription or a history of receiving a prescription, 21 U. S. C.
§ 353a(a), and the limitation on the percentage of a pharma-
cy’s total sales that out-of-state sales of compounded drugs
may represent, § 353a(b)(3)(B).

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
373 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
The Government has not offered any reason why these
possibilities, alone or in combination, would be insufficient
to prevent compounding from occurring on such a scale
as to undermine the new drug approval process. Indeed,
there is no hint that the Government even considered these
or any other alternatives. Nowhere in the legislative his-
tory of the FDAMA or petitioners’ briefs is there any ex-
planation of why the Government believed forbidding ad-
vertising was a necessary as opposed to merely convenient
means of achieving its interests. Yet “[i]t is well established
that ‘the party seeking to uphold a restriction on commercial
speech carries the burden of justifying it.’ ” Edenfield v.
Fane, 507 U. S., at 770 (quoting Bolger v. Youngs Drug Prod-
ucts Corp., 463 U. S. 60, 71, n. 20 (1983)). The Government
simply has not provided sufficient justification here. If the
First Amendment means anything, it means that regulating
speech must be a last—not first—resort. Yet here it seems
to have been the first strategy the Government thought
to try.
The dissent describes another governmental interest—an
interest in prohibiting the sale of compounded drugs to
“patients who may not clearly need them,” post, at 379 (opin-
ion of Breyer, J.)—and argues that “Congress could . . .
conclude that the advertising restrictions ‘directly advance’ ”
that interest, post, at 384. Nowhere in its briefs, however,
does the Government argue that this interest motivated
the advertising ban. Although, for the reasons given by the
dissent, Congress conceivably could have enacted the adver-
tising ban to advance this interest, we have generally only
sustained statutes on the basis of hypothesized justifications
when reviewing statutes merely to determine whether they
are rational. See L. Tribe, American Constitutional Law
1444–1446 (2d ed. 1988) (describing the “rational basis” or
“conceivable basis” test); see also, e. g., Minnesota v. Clover
Leaf Creamery Co., 449 U. S. 456, 466 (1981) (sustaining a
milk packaging regulation under the “rational basis” test

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
374 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
because “the Minnesota Legislature could rationally have
decided that [the regulation] might foster greater use of en-
vironmentally desirable alternatives” (emphasis deleted)).
The Central Hudson test is significantly stricter than the
rational basis test, however, requiring the Government not
only to identify specifically “a substantial interest to be
achieved by [the] restrictio[n] on commercial speech,” 447
U. S., at 564, but also to prove that the regulation “directly
advances” that interest and is “not more extensive than is
necessary to serve that interest,” id., at 566. The Govern-
ment has not met any of these requirements with regard to
the interest the dissent describes.
Even if the Government had argued that the FDAMA’s
speech-related restrictions were motivated by a fear that
advertising compounded drugs would put people who do
not need such drugs at risk by causing them to convince
their doctors to prescribe the drugs anyway, that fear would
fail to justify the restrictions. Aside from the fact that this
concern rests on the questionable assumption that doctors
would prescribe unnecessary medications (an assumption the
dissent is willing to make based on one magazine article and
one survey, post, at 383–384, neither of which was relied
upon by the Government), this concern amounts to a fear
that people would make bad decisions if given truthful
information about compounded drugs. See supra, at 368
(explaining that the Government does not claim the adver-
tisements forbidden by the FDAMA would be false or mis-
leading). We have previously rejected the notion that the
Government has an interest in preventing the dissemination
of truthful commercial information in order to prevent mem-
bers of the public from making bad decisions with the infor-
mation. In Virginia Bd. of Pharmacy, the State feared that
if people received price advertising from pharmacists, they
would “choose the low-cost, low-quality service and drive the
‘professional’ pharmacist out of business” and would “destroy
the pharmacist-customer relationship” by going from one

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
375 Cite as: 535 U. S. 357 (2002)
Opinion of the Court
pharmacist to another. We found these fears insufficient to
justify a ban on such advertising. 425 U. S., at 769. We
explained:
“There is, of course, an alternative to this highly pa-
ternalistic approach. That alternative is to assume that
this information is not in itself harmful, that people will
perceive their own best interests if only they are well
enough informed, and that the best means to that end is
to open the channels of communication rather than to
close them. . . . But the choice among these alternative
approaches is not ours to make or the Virginia General
Assembly’s. It is precisely this kind of choice, between
the dangers of suppressing information, and the dangers
of its misuse if it is freely available, that the First
Amendment makes for us. Virginia is free to require
whatever professional standards it wishes of its pharma-
cists; it may subsidize them or protect them from compe-
tition in other ways. . . . But it may not do so by keeping
the public in ignorance of the entirely lawful terms that
competing pharmacists are offering.” Id., at 770 (cita-
tion omitted).
See also 44 Liquormart, Inc. v. Rhode Island, 517 U. S., at
503 (“[B]ans against truthful, nonmisleading commercial
speech . . . usually rest solely on the offensive assumption
that the public will respond ‘irrationally’ to the truth. . . .
The First Amendment directs us to be especially skeptical
of regulations that seek to keep people in the dark for what
the government perceives to be their own good” (citation
omitted)).
Even if the Government had asserted an interest in pre-
venting people who do not need compounded drugs from
obtaining those drugs, the statute does not directly ad-
vance that interest. The dissent claims that the Govern-
ment “must exclude from the area of permitted drug sales . . .
those compounded drugs sought by patients who may not

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
376 THOMPSON v. WESTERN STATES MEDICAL CENTER
Opinion of the Court
clearly need them.” Post, at 379. Yet the statute does not
directly forbid such sales. It instead restricts advertising,
of course not just to those who do not need compounded
drugs, but also to individuals who do need compounded
drugs and their doctors. Although the advertising ban
may reduce the demand for compounded drugs from those
who do not need the drugs, it does nothing to prevent such
individuals from obtaining compounded drugs other than
requiring prescriptions. But if it is appropriate for the
statute to rely on doctors to refrain from prescribing com-
pounded drugs to patients who do not need them, it is not
clear why it would not also be appropriate to rely on doctors
to refrain from prescribing compounded drugs to patients
who do not need them in a world where advertising was
permitted.
The dissent may also be suggesting that the Government
has an interest in banning the advertising of compounded
drugs because patients who see such advertisements will be
confused about the drugs’ risks. See post, at 387 (“[The
Government] fears the systematic effect . . . of advertise-
ments that will not fully explain the complicated risks at
issue”). This argument is precluded, however, by the fact
that the Government does not argue that the advertisements
are misleading. Even if the Government did argue that it
had an interest in preventing misleading advertisements,
this interest could be satisfied by the far less restrictive al-
ternative of requiring each compounded drug to be labeled
with a warning that the drug had not undergone FDA test-
ing and that its risks were unknown.
If the Government’s failure to justify its decision to
regulate speech were not enough to convince us that the
FDAMA’s advertising provisions were unconstitutional, the
amount of beneficial speech prohibited by the FDAMA would
be. Forbidding the advertisement of compounded drugs
would affect pharmacists other than those interested in pro-
ducing drugs on a large scale. It would prevent pharmacists

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
377 Cite as: 535 U. S. 357 (2002)
Thomas, J., concurring
with no interest in mass-producing medications, but who
serve clienteles with special medical needs, from telling the
doctors treating those clients about the alternative drugs
available through compounding. For example, a pharmacist
serving a children’s hospital where many patients are un-
able to swallow pills would be prevented from telling the
children’s doctors about a new development in compounding
that allowed a drug that was previously available only in pill
form to be administered another way. Forbidding advertis-
ing of particular compounded drugs would also prohibit a
pharmacist from posting a notice informing customers that
if their children refuse to take medications because of the
taste, the pharmacist could change the flavor, and giving ex-
amples of medications where flavoring is possible. The fact
that the FDAMA would prohibit such seemingly useful
speech even though doing so does not appear to directly fur-
ther any asserted governmental objective confirms our belief
that the prohibition is unconstitutional.
Accordingly, we affirm the Court of Appeals’ judgment
that the speech-related provisions of FDAMA § 127(a) are
unconstitutional.
So ordered.
Justice Thomas, concurring.
I concur because I agree with the Court’s application of
the test set forth in Central Hudson Gas & Elec. Corp. v.
Public Serv. Comm’n of N. Y., 447 U. S. 557 (1980). I con-
tinue, however, to adhere to my view that cases such as this
should not be analyzed under the Central Hudson test.
“I do not believe that such a test should be applied to a re-
striction of ‘commercial’ speech, at least when, as here, the
asserted interest is one that is to be achieved through keep-
ing would-be recipients of the speech in the dark.” 44 Li-
quormart, Inc. v. Rhode Island, 517 U. S. 484, 523 (1996)
(opinion concurring in part and concurring in judgment).

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
378 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
Justice Breyer, with whom The Chief Justice, Jus-
tice Stevens, and Justice Ginsburg join, dissenting.
Federal law requires strict safety and efficacy testing of
all “new” prescription “drugs.” 21 U. S. C. § 355. See 21
CFR § 310.3(h) (2002) (defining “new drug” broadly). This
testing process requires for every “new drug” a preclinical
investigation and three separate clinical tests, including
small, controlled studies of healthy and diseased humans as
well as scientific double-blind studies designed to identify
any possible health risk or side effect associated with the
new drug. Practical Guide to Food and Drug Law and Reg-
ulation 95–102 (K. Pin˜ a & W. Pines eds. 1998). The objec-
tive of this elaborate and time-consuming regulatory regime
is to identify those health risks—both large and small—that
a doctor or pharmacist might not otherwise notice.
At the same time, the law exempts from its testing
requirements prescription drugs produced through “com-
pounding”—a process “by which a pharmacist or doctor com-
bines, mixes, or alters ingredients to create a medication
tailored to the needs of an individual patient.” Ante, at
360–361. The exemption is available, however, only if the
pharmacist meets certain specified conditions, including
the condition that the pharmacist not “advertise or promote
the compounding of any particular drug.” 21 U. S. C.
§ 353a(c) (emphasis added).
The Court holds that this condition restricts “commercial
speech” in violation of the First Amendment. See Central
Hudson Gas & Elec. Corp. v. Public Serv. Comm’n of N. Y.,
447 U. S. 557, 564 (1980). It concedes that the statutory pro-
vision tries to “[p]reserv[e] the effectiveness and integrity of
the . . . new drug approval process,” ante, at 369, and it
assumes without deciding that the statute might “ ‘directly
advance’ ” that interest, ante, at 371. It nonetheless finds
the statute unconstitutional because it could advance that
interest in other, less restrictive ways. Ante, at 372–373.
I disagree with this conclusion, and I believe that the Court

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
379 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
seriously undervalues the importance of the Government’s
interest in protecting the health and safety of the Ameri-
can public.
I
In my view, the advertising restriction “directly advances”
the statute’s important safety objective. That objective, as
the Court concedes, is to confine the sale of untested, com-
pounded, drugs to where they are medically needed. But to
do so the statute must exclude from the area of permitted
drug sales both (1) those drugs that traditional drug manu-
facturers might supply after testing—typically drugs capa-
ble of being produced in large amounts, and (2) those
compounded drugs sought by patients who may not clearly
need them—including compounded drugs produced in small
amounts.
The majority’s discussion focuses upon the first exclusion-
ary need, but it virtually ignores the second. It describes
the statute’s objective simply as drawing a “line” that will
“distinguish compounded drugs produced on such a small
scale that they could not undergo safety and efficacy testing
from drugs produced and sold on a large enough scale that
they could undergo such testing and therefore must do so.”
Ante, at 370 (emphasis added). This description overlooks
the need for a second line—a line that will distinguish
(1) sales of compounded drugs to those who clearly need
them from (2) sales of compounded drugs to those for whom
a specially tailored but untested drug is a convenience but
not a medical necessity. That is to say, the statute, in seek-
ing to confine distribution of untested tailored drugs, must
look both at the amount supplied (to help decide whether
ordinary manufacturers might provide a tested alternative)
and at the nature of demand (to help separate genuine need
from simple convenience). Cf. 143 Cong. Rec. S9840 (Sept.
24, 1997) (remarks of Sen. Kennedy) (understanding that
“some of the conditions are intended to ensure that the vol-
ume of compounding does not approach that ordinarily asso-

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
380 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
ciated with drug manufacturing” while others are “intended
to ensure that the compounded drugs that qualify for the
exemption have appropriate assurances of quality and safety
since [they] would not be subject to the more comprehensive
regulatory requirements that apply to manufactured drug
products”).
This second intermediate objective is logically related
to Congress’ primary end—the minimizing of safety risks.
The statute’s basic exemption from testing requirements
inherently creates risks simply by placing untested drugs
in the hands of the consumer. Where an individual has a
specific medical need for a specially tailored drug those risks
are likely offset. But where an untested drug is a conven-
ience, not a necessity, that offset is unlikely to be present.
That presumably is why neither the Food and Drug Ad-
ministration (FDA) nor Congress anywhere suggests that all
that matters is the total amount of a particular drug’s sales.
That is why the statute’s history suggests that the amount
supplied is not the whole story. See S. Rep. No. 105–43,
p. 67 (1997) (statute seeks to assure “continued availability of
compounded drug products as a component of individualized
therapy, . . . while . . . prevent[ing] small-scale manufac-
turing under the guise of compounding” (emphasis added));
accord, H. R. Conf. Rep. No. 105–399, p. 94 (1997). That is
why the statute itself, as well as the FDA policy that the
statute reflects, lists several distinguishing factors, of which
advertising is one. See FDA Compliance Policy Guide
7132.16, reprinted in App. to Pet. for Cert. 71a–77a (here-
inafter Compliance Policy Guide). And that is likely why,
when faced with the possibility of severing the advertising
restriction from the rest of the statute, the Government ar-
gued that the “other conditions in section 353a alone are in-
adequate to achieve Congress’s desired balance among com-
peting interests.” See Brief for Appellants in No. 99–17424
(CA9), p. 57. See also id., at 55 (to nullify advertising re-
strictions would undermine “ ‘finely tuned balance’ ” achieved

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
381 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
by requiring that “pharmacies refrain from promoting and
soliciting prescriptions for particular compounded drug prod-
ucts until they have been proven safe and effective”).
Ensuring that the risks associated with compounded drug
prescriptions are offset by the benefits is also why public
health authorities, testifying in Congress, insisted that the
doctor’s prescription represent an individualized determi-
nation of need. See, e. g., FDA Reform Legislation: Hear-
ings before the Subcommittee on Health and the Environ-
ment of the House Committee on Commerce, 104th Cong.,
2d Sess., 120 (1996) (hereinafter FDA Reform Legislation)
(statement of Mary K. Pendergast, Deputy Commissioner of
the FDA and Senior Advisor to the Commissioner) (Allowing
traditional compounding is “good medicine” because “an indi-
vidual physician” was making “an individualized determina-
tion for a patient”). See also National Association of Boards
of Pharmacy, Model State Pharmacy Act and Rules, Art. I,
§ 1.05(e) (1996) (hereinafter NABP Model Act) (defining
“[c]ompounding” as involving a prescription “based on the
Practitioner/patient/Pharmacist relationship in the course of
professional practice”).
And that, in part, is why federal and state authorities have
long permitted pharmacists to advertise the fact that they
compound drugs, while forbidding the advertisement of indi-
vidual compounds. See Compliance Policy Guide 76a; Good
Compounding Practices Applicable to State Licensed Phar-
macies, NABP Model Act, App. C.2, subpart A (forbidding
pharmacists to “solicit business (e. g., promote, advertise,
or use salespersons) to compound specific drug products”).
The definitions of drug manufacturing and compounding
used by the NABP and at least 13 States reflect similar dis-
tinctions. NABP Model Act, Art. I, §§ 105(e), (t), and (u)
(defining drug manufacturing to “include the promotion and
marketing of such drugs or devices” but excluding any refer-
ence to promotion or marketing from the definition of drug
compounding); Alaska Stat. §§ 08.80.480(3) and (15) (2000)

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
382 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
(same); La. Stat. Ann. §§ 37:1164(5) and (25) (West 2000)
(same); Miss. Code Ann. §§ 73–21–73(c) and (s) (Lexis 1973–
2000) (same); Mont. Code Ann. § 37–7–101(7) (1997) (same);
N. H. Rev. Stat. Ann. §§ 318–1(III) and (VIII) (Supp. 2001)
(same); N. M. Stat. Ann. §§ 61–11–2(C) and (Q) (2001) (same);
Ohio Rev. Code Ann. § 3715.01(14) (West Supp. 2002) (same);
Okla. Stat., Tit. 59, §§ 353.1(20) and (26) (Supp. 2002) (same);
S. C. Code Ann. §§ 40–43–30(7) and (29) (2001) (same); Tenn.
Code Ann. §§ 63–10–404(4) and (18) (1997) (same); Tex. Occ.
Code Ann. §§ 551.003(9) and (23) (2002 Pamphlet) (same);
W. Va. Code §§ 30–5–1b(c) and (o) (1966–1998) (same).
These policies and statutory provisions reflect the view
that individualized consideration is more likely present,
and convenience alone is more likely absent, when demand
for a compounding prescription originates with a doctor,
not an advertisement. The restrictions try to assure that
demand is generated doctor-to-patient-to-pharmacist, not
pharmacist-to-advertisement-to-patient-to-doctor. And they
do so in order to diminish the likelihood that those who do
not genuinely need untested compounded drugs will not re-
ceive them.
There is considerable evidence that the relevant means—
the advertising restrictions—directly advance this statu-
tory objective. No one denies that the FDA’s complex test-
ing system for new drugs—a system that typically relies
upon double-blind or other scientific studies—is more likely
to find, and to assess, small safety risks than are physicians
or pharmacists relying upon impressions and anecdotes.
See supra, at 378.
Nor can anyone deny that compounded drugs carry with
them special risks. After all, compounding is not neces-
sarily a matter of changing a drug’s flavor, cf. ante, at 377,
but rather it is a matter of combining different ingredients
in new, untested ways, say, adding a pain medication to an
antihistamine to counteract allergies or increasing the ratio
of approved ingredients in a salve to help the body absorb it

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
383 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
at a faster rate. And the risks associated with the untested
combination of ingredients or the quicker absorption rate or
the working conditions necessary to change an old drug into
its new form can, for some patients, mean infection, serious
side effects, or even death. See, e. g., J. Thompson, A Practi-
cal Guide to Contemporary Pharmacy Practice 11.5 (1998)
(hereinafter Contemporary Pharmacy Practice). Cf. 21
CFR § 310.3(h)(1) (2002) (considering a drug to be “new” and
subject to the approval process if the “substance which com-
poses such drug” is new); § 310.3(h)(3) (considering a drug to
be “new” and subject to the approval process if approved
ingredients are combined in new proportions).
There is considerable evidence that consumer oriented ad-
vertising will create strong consumer-driven demand for a
particular drug. See, e. g., National Institute for Health
Care Management, Factors Affecting the Growth of Pre-
scription Drug Expenditures iii (July 9, 1999) (three anti-
histamine manufacturers spent $313 million on advertising
in 1998 and accounted for 90% of prescription drug anti-
histamine market); Kritz, Ask Your Doctor About . . . Which
of the Many Advertised Allergy Drugs Are Right for You?
Washington Post, June 6, 2000, Health, p. 9 (The manu-
facturer of the world’s top selling allergy drug, the eighth
best-selling drug in the United States, spent almost $140
million in 1999 on advertising); 1999 Prevention Magazine 10
(spending on direct-to-consumer advertising of prescription
medicine increased from $965.2 million in 1997 to $1.33 billion
in 1998).
And there is strong evidence that doctors will often re-
spond affirmatively to a patient’s request for a specific drug
that the patient has seen advertised. See id., at 32 (84%
of consumers polled report that doctors accommodate their
request for a specific drug); Henry J. Kaiser Family Founda-
tion, Understanding the Effects of Direct-to-Consumer Pre-
scription Drug Advertising 3 (Nov. 2001) (A foundation sur-
vey found that more than one in eight Americans had asked

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
384 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
for—and received—a specific prescription from their doctor
in response to an advertisement).
In these circumstances, Congress could reasonably con-
clude that doctors will respond affirmatively to a patient’s
request for a compounded drug even if the doctor would not
normally prescribe it. When a parent learns that a child’s
pill can be administered in liquid form, when a patient learns
that a compounded skin cream has an enhanced penetration
rate, or when an allergy sufferer learns that a compounded
antiinflammatory/allergy medication can alleviate a sinus
headache without the sedative effects of antihistamines, that
parent or patient may well ask for the desired prescription.
And the doctor may well write the prescription even in the
absence of special need—at least if any risk likely to arise
from lack of testing is so small that only scientific testing,
not anecdote or experience, would reveal it. It is conse-
quently not surprising that 71% of the active members of
the American Academy of Family Physicians “believe that
direct-to-consumer advertising pressures physicians into
prescribing drugs that they would not ordinarily prescribe.”
Rosenthal, Berndt, Donohue, Frank, & Epstein, Promotion
of Prescription Drugs to Consumers, 346 New Eng. J. Med.
498–505 (2002) (citing Lipsky, The Opinions and Experiences
of Family Physicians Regarding Direct-To-Consumer Adver-
tising, 45 J. Fam. Pract. 495–499 (1997)).
Of course, the added risks in any such individual case may
be small. But those individual risks added together can
significantly affect the public health. At least, the FDA and
Congress could reasonably reach that conclusion. And that
fact, along with the absence of any significant evidence that
the advertising restrictions have prevented doctors from
learning about, or obtaining, compounded drugs, means that
the FDA and Congress could also conclude that the advertis-
ing restrictions “directly advance” the statute’s safety goal.
They help to assure that demand for an untested com-
pounded drug originates with the doctor, responding to an

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
385 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
individual’s special medical needs; they thereby help to re-
strict the untested drug’s distribution to those most likely to
need it; and they thereby advance the statute’s safety goals.
There is no reason for this Court, as a matter of constitu-
tional law, to reach a different conclusion.
II
I do not believe that Congress could have achieved its
safety objectives in significantly less restrictive ways. Con-
sider the several alternatives the Court suggests. First, it
says that “the Government could ban the use of ‘commercial
scale manufacturing or testing equipment for compounding
drug products.’ ” Ante, at 372. This alternative simply
restricts compounding to drugs produced in small batches.
It would neither limit the total quantity of compounded
drugs produced, nor help in any way to assure the kind of
individualized doctor-patient need determination that the
statute’s advertising restriction are designed to help achieve.
Second, the Court says that the Government “could pro-
hibit pharmacists from compounding more drugs in anticipa-
tion of receiving prescriptions than in response to pre-
scriptions already received.” Ibid. This alternative, while
addressing the issue of quantity, does virtually nothing to
promote the second, need-related statutory objective.
Third, the Court says the Government “could prohibit
pharmacists from ‘[o]ffering compounded drug products at
wholesale to other state licensed persons or commercial
entities for resale.” Ibid. This alternative is open to the
same objection.
Fourth, the Court says the Government “could limit the
amount of compounded drugs, either by volume or by num-
bers of prescriptions, that a given pharmacist or pharmacy
sells out of state.” Ibid. This alternative, applying only
to out-of-state sales, would not significantly restrict sales,
either in respect to amounts or in respect to patient need.

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
386 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
In fact, it could prevent compounded drugs from reaching
out-of-state patients who genuinely need them.
Fifth, the Court says that the Government could “ca[p] the
amount of any particular compounded drug, either by drug
volume, number of prescriptions, gross revenue, or profit.”
Ibid. This alternative, like the others, ignores the patient-
need problem, while simultaneously threatening to prevent
compounded drugs from reaching those who genuinely need
them, say, a patient whose prescription represents one be-
yond the arbitrarily imposed quantitative limit.
Sixth, the Court says that the Government could rely
upon “non-speech-related provisions of the FDAMA, such
as the requirement that compounding only be conducted in
response to a prescription.” Ibid. This alternative also
ignores the patient-need problem and was specifically re-
jected by the Government in the Court of Appeals for the
Ninth Circuit. See supra, at 380–381.
The Court adds that “[t]he Government has not offered
any reason why these possibilities, alone or in combination,
would be insufficient.” Ante, at 373. The Government’s
failure to do so may reflect the fact that only the Court,
not any of the respondents, has here suggested that these
“alternatives,” alone or in combination, would prove suffi-
cient. In fact, the FDA’s Compliance Policy Guide, from
which the Court draws its first four alternatives, specifi-
cally warned that these alternatives alone were insuffi-
cient to successfully distinguish traditional compounding
from unacceptable manufacturing. See Compliance Policy
Guide 77a.
III
The Court responds to the claim that advertising com-
pounded drugs causes people to obtain drugs that do not
promote their health, by finding it implausible given the
need for a prescription and by suggesting that it is not rele-
vant. The First Amendment, it says, does not permit the
Government to control the content of advertising, where

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
387 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
doing so flows from “fear” that “people would make bad
decisions if given truthful information about compounded
drugs.” Ante, at 374. This response, however, does not
fully explain the Government’s regulatory rationale; it fails
to take account of considerations that make the claim more
than plausible (if properly stated); and it is inconsistent with
this Court’s interpretation of the Constitution.
It is an oversimplification to say that the Government
“fear[s]” that doctors or patients “would make bad deci-
sions if given truthful information.” Ibid. Rather, the
Government fears the safety consequences of multiple
compound-drug prescription decisions initiated not by doc-
tors but by pharmacist-to-patient advertising. Those con-
sequences flow from the adverse cumulative effects of multi-
ple individual decisions each of which may seem perfectly
reasonable considered on its own. The Government fears
that, taken together, these apparently rational individual de-
cisions will undermine the safety testing system, thereby
producing overall a net balance of harm. See, e. g., FDA
Reform Legislation 121 (statement of David A. Kessler,
Commissioner of the FDA) (voicing concerns about “quality
controls” and the integrity of the drug-testing system).
Consequently, the Government leaves pharmacists free to
explain through advertisements what compounding is, to ad-
vertise that they engage in compounding, and to advise pa-
tients to discuss the matter with their physicians. And it
forbids advertising the specific drug in question, not because
it fears the “information” the advertisement provides, but
because it fears the systematic effect, insofar as advertise-
ments solicit business, of advertisements that will not fully
explain the complicated risks at issue. And this latter fear
is more than plausible. See Part I, supra.
I do not deny that the statute restricts the circulation of
some truthful information. It prevents a pharmacist from
including in an advertisement the information that “this
pharmacy will compound Drug X.” Nonetheless, this Court

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
388 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
has not previously held that commercial advertising restric-
tions automatically violate the First Amendment. Rather,
the Court has applied a more flexible test. It has examined
the restriction’s proportionality, the relation between re-
striction and objective, the fit between ends and means. In
doing so, the Court has asked whether the regulation of com-
mercial speech “directly advances” a “substantial” govern-
mental objective and whether it is “more extensive than is
necessary” to achieve those ends. See Central Hudson, 447
U. S., at 566. It has done so because it has concluded that,
from a constitutional perspective, commercial speech does
not warrant application of the Court’s strictest speech-
protective tests. And it has reached this conclusion in part
because restrictions on commercial speech do not often re-
press individual self-expression; they rarely interfere with
the functioning of democratic political processes; and they
often reflect a democratically determined governmental deci-
sion to regulate a commercial venture in order to protect,
for example, the consumer, the public health, individual
safety, or the environment. See, e. g., 44 Liquormart, Inc.
v. Rhode Island, 517 U. S. 484, 499 (1996) (“[T]he State’s
power to regulate commercial transactions justifies its con-
comitant power to regulate commercial speech that is ‘linked
inextricably’ to those transactions”); L. Tribe, American
Constitutional Law § 12–15, p. 903 (2d ed. 1988) (“[C]ommer-
cial speech doctrine” seeks to accommodate “the right to
speak and hear expression about goods and services” with
“the right of government to regulate the sales of such goods
and services” (emphasis in original)).
I have explained why I believe the statute satisfies this
more flexible test. See Parts I and II, supra. The Court,
in my view, gives insufficient weight to the Government’s
regulatory rationale, and too readily assumes the existence
of practical alternatives. It thereby applies the commercial
speech doctrine too strictly. Cf. Buckman Co. v. Plaintiffs’
Legal Comm., 531 U. S. 341, 349 (2001) (flexibility necessary

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
389 Cite as: 535 U. S. 357 (2002)
Breyer, J., dissenting
if FDA is to “pursu[e] difficult (and often competing) objec-
tives”). See also Illinois Bd. of Elections v. Socialist Work-
ers Party, 440 U. S. 173, 188–189 (1979) (Blackmun, J., con-
curring) (warning against overly demanding search for less
restrictive alternatives).
In my view, the Constitution demands a more lenient
application, an application that reflects the need for dis-
tinctions among contexts, forms of regulation, and forms
of speech, and which, in particular, clearly distinguishes
between “commercial speech” and other forms of speech
demanding stricter constitutional protection. Otherwise, an
overly rigid “commercial speech” doctrine will transform
what ought to be a legislative or regulatory decision about
the best way to protect the health and safety of the Ameri-
can public into a constitutional decision prohibiting the legis-
lature from enacting necessary protections. As history in
respect to the Due Process Clause shows, any such trans-
formation would involve a tragic constitutional misunder-
standing. See id., at 189 (Blackmun, J., concurring).
IV
Finally, the majority would hold the statute unconsti-
tutional because it prohibits pharmacists from advertising
compounded drugs to doctors. Ante, at 376–377. Doctors,
however, obtain information about individual drugs through
many other channels. And there is no indication that re-
strictions on commercial advertising have had any negative
effect on the flow of this information. See e. g., Contem-
porary Pharmacy Practice 11.4 (compounded drug informa-
tion “available” and “widely disseminated” through books,
journals, monographs, and vendors). Nor, with one excep-
tion, have doctors or groups of doctors complained that the
statute will interfere with that flow of information in the
future. But see Brief for Julian M. Whitaker, M.D., et al.
as Amici Curiae 1 (alleging, without evidentiary support,
that the regulations prevent doctors from knowing how to

535US2 Unit: $U41 [09-24-03 18:37:54] PAGES PGT: OPIN
390 THOMPSON v. WESTERN STATES MEDICAL CENTER
Breyer, J., dissenting
get “competitively priced compounded drugs as efficiently
as possble”).
Regardless, we here consider a facial attack on the statute.
The respondents here focus their attack almost entirely
upon consumer-directed advertising. They have not fully
addressed separate questions involving the effect of adver-
tising restrictions on information received by physicians.
I would consequently leave these questions in abeyance.
Considering the statute only insofar as it applies to advertis-
ing directed at consumers, I would hold it constitutional.
For these reasons, I dissent.