This case was instituted under the Federal Food, Drug, and Cosmetic Act, 21 U.S.C. § 301 et seq., for seizure and condemnation of a quantity of drug called “Afrodex” marketed by Bentex Pharmaceutical Company, claimant below. The libel alleges that Afrodex is misbranded under 21 U.S.C. § 352(f)(1), and thus subject to seizure and condemnation pursuant to 21 U.S.C. § 334, in that the article is, within the meaning of 21 U.S.C. § 321 (p), a “new drug” whose labeling is not specified in an approved New Drug Application (NDA). The parties stipulate that if Afrodex is a “new drug” it is misbrand-ed within the meaning of 21 U.S.C. § 352(f) (1). '
I. The statutory scheme
The Federal Food, Drug and Cosmetic Act, passed in 1938 and amended in 1962, defines a “new drug” as follows:
The term “new drug” means — (1) Any drug . . . the composition of which is such that such drug is not generally recognized, among experts qualified by scientific training and experience to evaluate the safety and effectiveness of drugs, as safe and effective for use under the conditions prescribed, recommended, or suggested in the labeling thereof ....
21 U.S.C. § 321(p). The italicized language was added by the Drug Amendments Act of 1962. Thus under present standards a “new drug” is one not generally recognized by qualified experts as both safe and effective under the conditions of use prescribed, recommended or *1160suggested in its labeling. The 1962 amendments, however, also contained a “grandfather clause” relieving old drugs then on the market from the “effectiveness” provision of the definition of a “new drug” under certain circumstances.1 Under that clause a product is not a “new drug” even if it is not generally recognized as effective if it was marketed on October 9, 1962, was then generally recognized as safe, was not then covered by an effective NDA, and is still intended for use solely under the conditions prescribed as of October 9, 1962.
Under the scheme of the Act the ultimate determination of the safety of a drug is not a matter given to the courts, but one to be determined by the Food & Drug Administration upon submission of an NDA. Thus a decision adverse to claimant does not mean that Afrodex is unsafe or ineffective but simply that it has not been afforded that general recognition among qualified experts which when demonstrated exempts a drug from the NDA requirements of the Act.
II. The issues
After a complex trial the District Judge in an opinion reported at 347 F. Supp. 768 determined that (1) Afrodex was not a “grandfathered” drug because on October 9, 1962, it was not generally recognized among qualified experts as safe under the conditions of use prescribed, recommended, or suggested in its labeling; and (2) Afrodex is not presently generally recognized among qualified experts as safe and effective for the conditions of -use prescribed, recommended, or suggested in its labeling and is thus a “new drug.”
Bentex’s attack on the findings and conclusions of the trial court are directed at what it terms the evaluation of the evidence by the District Judge, in four respects. (A.) The safety of Afrodex should be judged by the safety of its three constituent ingredients and thus the District Judge should not have considered the lack of evidence of safety in combination as probative of a lack of general recognition of safety. (B.) In evaluating the general recognition of safety of the individual ingredients the District Judge relied on evidence that the ingredients were unsafe when taken in massive doses while the proper standard is safety at recommended or prescribed dosage. (C.) In evaluating the general recognition of safety of the individual ingredients the District Court erred by considering hazards resulting from a use contraindicated on the label. (D.) In evaluating whether Afrodex is presently generally recognized as effective for its intended uses the District Judge applied meanings to uses indicated other than those intended by the label.
A. Safety in combination
Bentex asserts as reversible error this statement in the opinion of the District Court:
It should be noted that the safety and efficacy of combination drugs such as Afrodex cannot be equated with the safety of the components separately or in combination with different ingredients, United States v. Seven Cartons, More or Less, etc., 293 F.Supp. 660, 664 (S.D.Ill.1968). Thus, the fact that any one individual component of a combination drug may be generally recognized as safe and effective is not *1161relevant to the issue asserted, that is, whether the combination itself is so recognized.
Claimant’s first premise is that whether safety and effectiveness in combination may be determined from individual ingredient safety and effectiveness is a question to be determined in each case from the testimony of experts rather than a question to be resolved by application of a legal rule. Assuming this premise to be correct,2 however, would not afford a sufficient basis for reversal. Read in the context of the opinion as a whole, the statement represents but an alternative ground, and a minor one at that, on which the District Court based its conclusion that Afrodex is not and has never been generally recognized as safe. The paragraph in which the statement occurs begins,
A major portion of the expert testimony heard in this suit centered around the safety and efficacy of the component methyltestosterone. Expert testimony was also taken concerning yohimbine and extract of nux vomica, as well as the combination of these ingredients in the amounts contained in the drug Afrodex.
The clearly evident emphasis is not on the combination but on the ingredients individually. Similarly the District Court’s summary of the testimony in the paragraphs following this one contains extensive references to the individual ingredients and comparatively little reference to safety in combination. Read in context, the statement on which Bentex would predicate reversal is little more than a notation in passing. The major ground for the District Court’s conclusion was the determination that the individual ingredients of Afrodex were not generally recognized as safe. The focus of the remaining inquiries is on the general recognition of safety of the individual ingredients.
B. Dosage
We reject Bentex’s argument that the testimony of the government’s expert witnesses concerning lack of safety of the individual ingredients related to massive dosages rather than the dosage (20 milligrams of each active ingredient per day) prescribed on the label. Dr. Arthur Grollman, a professor of medicine appearing as an expert witness for the government, was asked whether in his opinion the ingredients in Afrodex “both singly and in combination” were safe “for treatment of the conditions for which Afrodex is offered,” and responded that under certain conditions use of those ingredients might not be safe. It is evident from Dr. Grollman’s prior testimony that he knew the dosage of each of the individual ingredients contained in Afrodex.3 The District Court was entitled to infer that a competent expert medical witness, knowing the composition of a drug, when asked about the safety of the ingredients of that drug for use in the conditions specified, would respond in terms of the lack of safety at the dosage prescribed. Bentex urges that such an inference is un*1162supportable because in cross-examination its counsel elicited several concessions damaging to the government regarding the probable small dosage safety of two of Afrodex’s ingredients, nux vomica and yohimbine. But the focus of the testimony concerning safety, as distinguished from effectiveness, was not on these ingredients but rather on the ingredient methyltestosterone. Dr. Emil Steinberger, a professor of medicine specializing in male reproductive biology, testified that liver inflammation and jaundice are possible side effects of methyltestosterone.4 While Dr. Stein-berger’s testimony did not directly state the dosage necessary to make such effects likely, he was aware of the dosage contained in Afrodex and, in addition, testified that liver disease was not but should have been listed on the labeling of Afrodex as a contraindication. These factors adequately demonstrate a connection between his testimony and the dosage contained in Afrodex. In addition to the evidence concerning liver damage there was specific testimony that a dosage of methyltestosterone even smaller than that contained in Afrodex could activate latent cancer of the prostate.5
C. Contraindication in cases of “known or suspected malignancies”
The testimony, not disputed by claimant’s witnesses, that administration of methyltestosterone may activate latent cancer of the prostate, brings us to Ben-tex’s third claimed error. Bentex objects that the danger of such activation may not properly be considered in determining whether Afrodex is “generally recognized ... as safe for use under the. conditions prescribed, recommended, or suggested in the labeling thereof . . . ” since the label contains a contraindication of use in cases of “known or suspected malignancies.” But Dr. Scott’s testimony was to the effect that in four out of five cases a patient may have latent cancer of the prostate though not “known or suspected.” Consequently use of the drug in circumstances allowed by the label may create the danger of activation.
Claimant also suggests that the risk of activation is not limited to methyltestosterone but results equally from the administration of any similar androgen replacement and that administration of such replacements, including methyl-testosterone, is recognized as useful and proper therapy in cases of androgen deficiency and has in fact been so recognized through notice in the Federal Register by the FDA itself. The relevance of this attack to the issue of general recognition of safety is questionable,6 but we need not dismiss it on that basis. Testimony by numerous government witnesses, notably Dr. Schoolar, indicated *1163that the vast majority of cases of impotence were psychogenic in origin and that in such cases androgen replacement therapy through administration of meth-yltestosterone was not a beneficial treatment since there was no androgen deficiency to begin with. The District Court specifically found that the indicated uses of Afrodex were not limited to cases of • androgenic deficiency.7 Dr. Grollman under cross-examination by the claimant’s counsel testified that a general practitioner would not read the labeling of Afrodex as limiting its recommended uses to cases of androgen deficiency. His testimony was borne out by testimony of several practicing physicians appearing as witnesses for the claimant who testified that they believed Afrodex to be indicated for treatment of impotence unrelated to androgen deficiency and, in at least one instance, believed it to be indicated for neurasthenia unaccompanied by impotence. See 347 F.Supp. at 775-776. There is then testimony that Afrodex is a drug which, according to claimant’s witnesses, is indicated for use in circumstances where, according to the government’s witnesses, it has no therapeutic value. Also there is testimony that its use generates a risk of activating a latent carcinoma which is common, difficult to detect, and typically undetected.8 This combined testimony undergirds the conclusion drawn by the government’s witnesses and the District Court that administration of Afrodex, which includes a meth-yltestosterone component, is not now and never has been “generally recognized among experts qualified by scientific training and experience to evaluate the safety of drugs as safe for use under the conditions prescribed, recommended or suggested in the labeling thereof if
D. The meaning of the label
Bentex attempts through a rather artificial construction of the wording of the label to limit the label’s recommended uses for Afrodex to cases involving androgen deficiency and asserts that the District Court erred by “applying meanings to the uses indicated other than those intended by the label.” 9 No matter what Bentex or the inanimate label may have intended, we *1164have already noted that medical witnesses on both sides read the label as recommending use in cases beyond those involving androgen deficiency. Their testimony provided ample support for the meanings found by the District Court.
•X- *x* *x-
Claimant’s arguments, with the possible exception of the first, are without merit. And even assuming that claimant is correct in its first contention, the decision of the District Court was amply justified on the ground that at least one of the individual ingredients of Afrodex is not now and was not in 1962 generally recognized as safe for use in the conditions recommended on the labeling of Afrodex.
Affirmed.
. The “grandfather clause” provides:
In the case of any drug which, on the day immediately preceding the enactment date [October 10, 1962], (A) was commercially used or sold in the United States, (B) was not a new drug as defined by section 201 (p) [21 U.S.C. § 321(p) ] of the basic Act as then in force, and (C) was not covered by an effective application under section 505 [21 U.S.C. § 355] of that Act, the amendments to section 201 (p) [21 U. S.C. § 321 (p) ] made by this Act shall not apply to such drug when intended solely for use under conditions prescribed, recommended, or suggested in labeling with respect to such drug on that day.
Footnote to 21 U.S.C.A. § 321.
. The government argues that we have previously determined this premise to be unfounded and the statement of the District Court to be correct by our affirmance “for the reasons set out in the district court’s opinion” in United States v. Article of Drug . Furestrol Vaginal Suppositories, 415 F.2d 390 (CA5, 1969), affirming 294 F. Supp. 1307 (N.D.Ga.1968). In that case the drug in question was composed of two active ingredients, both of which were individually generally recognized as safe and effective for certain uses. One of these ingredients, however, was not generally recognized as safe and effective for the uses specified on the labeling of the combination drug, and testimony concerning the safety and efficacy of this ingredient in the labeled use formed the basis for concluding that the combination was a “new drug.” Thus, contrary to the government’s contention, the case does not support the proposition that as a matter of law general recognition of safety of a combination drug in its labeled uses may not in any case be demonstrated from general recognition of safety of its constituent ingredients in those same uses.
. Similarly the testimony of the other government witnesses reveals that each was aware of the dosage.
. This testimony was corroborated by the testimony of Dr. Joseph Schoolar, a professor of pharmacology and psychiatry, who testified that liver disease was a generally recognized side effect of methyltestosterone.
. The attack’s relevance is questionable because its basis may be that even though administration of methyltestosterone carries with it a risk, taking that risk may be justified in certain instances because of the benefits expected. Thus even if methyltestos-terone is recognized as a useful and proper therapy in some cases, it does not follow that it is generally recognized by qualified *1163experts as safe. To conclude otherwise would put courts into the business of making sophisticated risk versus benefit evaluations of actual safety better left to the FDA. Gf. Weinberger v. Bentex Pharmaceuticals, Inc., 412 U.S. 645, 93 S.Ct. 2488, 37 L.Ed.2d 235, 242-243 (1973).
. The label of Afrodex recommends its use in the conditions of “hypogonadal impotence and/or male climacteric; neurasthenia.” The District Court made the following findings with regard to the meaning of these terms.
“Hypogonadal impotence is impotence caused by insufficient functioning of the testes to secrete testosterone, that is, impotence caused by androgen deficiency. Male climacteric is a somewhat outdated, decidedly controversial term used variously to describe (1) an organically caused condition in the male analogous to menopause in the female brought about by a sudden drop of androgen production, or (2) a condition systemic or psychogenic in origin characterized by hot flashes, irritability and impotence. In this sense, the term male climacteric does not indicate a decrease in androgen production. Neurasthenia, as used in the labeling of Afrodex, is an archaic, obsolete term used to describe a general condition of weakness or nervous fatigue. Testimony reflects that neurasthenia refers to symptoms rather than to specific diseases or malfunctions. The condition of neurasthenia is viewed as psychogenic in origin.”
. Thus even if we assume that the weighing of risks versus benefits in this case is to some extent a task for courts rather than the FDA, see note 6, supra, balancing the asserted risk of activation against the asserted lack of benefit in cases not involving androgen deficiency would require the conclusion that Afrodex is not and never has been generally recognized as safe for its labeled uses.
. The relevant wording and the District Court’s findings as to its meaning are quoted at note 7, supra. Bentex raised this objection in regard to the finding of a lack of general recognition of effectiveness, an issue which we find it unnecessary to address. However, since we have determined that the meaning of the label is relevant to general recognition of safety we believe it appropriate to address the objection.