Opinion for the Court filed by Circuit Judge GINSBURG.
GINSBURG, Circuit Judge:Nature has well equipped the flea to meet the life-or-death challenge of alighting upon a suitable host; it is estimated that if humans possessed a jumping ability comparable to that of Ctenocephalides cams1 a person could leap five city blocks in a single bound. IV Encyclopedia Britannica 177 (15th ed. 1974); see also M. Rothchild & T. Clay, Fleas, Flukes and Cuckoos 42 (1952). Analogously, Congress in 1968 equipped the Food and Drug Administration (FDA or Administration) with a consolidated statutory framework to meet the challenge of regulating new animal drugs. See Animal Drug Amendments of 1968, Pub.L. No. 90-399, 82 Stat. 343 (codified as 21 U.S.C. § 360b (1982)); S.Rep. No. 1308, 90th Cong., 2d Sess. 1-3 (1968), U.S.Code Cong. & Admin.News 1968, p. 2607. The petitioner in this case, who seeks to market over the counter (OTC) a flea-control product for dogs, asserts that Congress did not equip the FDA as generously as nature did the insectile subject of this proceeding. Specifically, petitioner argues that because the FDA seventeen years ago approved petitioner’s product as safe and effective for distribution by veterinarian’s prescription, the Administration is powerless to reevaluate the product’s safety and effectiveness before allowing the product to be marketed over the counter.
We reject petitioner’s argument. As we read the governing statute, the FDA may fully evaluate the safety and effectiveness of a product not only before approving an initial new animal drug application (NADA), but also before approving a supplemental NADA when the proposed change (such as prescription-only to' OTC status) has a bearing on the product’s safety or effectiveness.
Having found the FDA’s investigation within its statutory authority, we further find the agency’s conclusion- — that petitioner’s product has not been proven safe and effective — to be supported by substantial evidence and not otherwise arbitrary or capricious. We therefore affirm the FDA’s ruling that petitioner’s supplemental NADA seeking OTC status was not up to scratch.
I. Background
In the mid-1960’s, petitioner, American Cyanamid Company (Cyanamid), filed two new drug applications with the FDA, one for the liquid form, and the other for the capsule form of Proban, a drug intended to control fleas on dogs.2 In a manner reminiscent of the lifecycle of Proban’s intended victim,3 the active ingredient of Proban *302travels an indirect route from bottle to parasite. The drug is placed in a dog’s food; the dog ingests the drug; a flea ingests the dog's body fluids; the flea eventually ingests Proban from the canine blood and, it is hoped, dies.
In 1968, the FDA approved the liquid and capsule forms of Proban as safe and effective for distribution under the order of a licensed veterinarian.
In 1972, Cyanamid filed a supplemental NADA seeking authorization to market liquid Proban over the counter. Working at a pace charitably described as unhurried,4 the Administration took two years to inform Cyanamid by letter that its application was “incomplete,” then two more years to issue a Notice of Opportunity for Hearing, and ultimately, one more year to issue a final order rejecting the supplemental NADA and refusing Cyanamid's request for a hearing. American Cyanamid Co. v. Food and Drug Administration, 606 F.2d 1307, 1309 & n. 4 (D.C.Cir.1979).
Cyanamid petitioned this court to review the Administration’s ruling. The appellate panel reversed the agency’s order and remanded the case to the FDA, directing that Cyanamid receive the trial-type hearing it had requested. The FDA had invoked its summary disposition mechanism, allowing denial of an application without a hearing if the application presents “no genuine and substantial issue of fact.” See 21 C.F.R. § 514.200(c) (1978). This court “recognized a rather broad area” in which the FDA may properly pluck out meritless applications, American Cyanamid, 606 F.2d at 1323; inspecting Cyanamid’s papers, however, the court discovered “several material issues of fact and science” entitling the applicant to a hearing. Id.
Cyanamid’s victory proved fleeting. In 1980 the Administration published a Notice detailing the factual issues to be considered at Cyanamid’s hearing. The list included specific questions concerning the adequacy of Cyanamid’s data on the safety and effectiveness of Proban. 45 Fed.Reg. 40236 (1980). A prehearing conference identified six main issues for determination:
I. Has Cyanamid submitted tests by all methods reasonably applicable to show that Proban is safe for OTC use?
II. Do the results of the tests show that the proposed use of Proban is safe?
III. Upon the basis of the information submitted by Cyanamid or other information before the Commissioner, does the Commissioner have sufficient information to determine whether the proposed use of Proban is safe?
IV. Has Cyanamid submitted substantial evidence, consisting of adequate and well-controlled investigations, including field investigations by experts qualified by scientific training and experience to evaluate the effectiveness of Proban, to establish that Proban marketed OTC will have the effect it purports or is represented to have under its prescribed, recommended, or suggested conditions of use?
V. Whether the applicant has shown by appropriate studies, data, surveys or opinion evidence that the proposed labeling for Proban constitutes adequate directions for use?
VI. Are the dangers associated with Proban such that the proposed labeling *303for Proban does not provide instructions for lay use?
Proposal to Refuse Approval of a Supplemental New Animal Drug Application for Over-the-Counter Sale of Proban Brand of Cythioate Oral Liquid, Initial Decision, No. 76N-0462 (June 16, 1981) [hereinafter Initial Decision] at 3-4. These issues were drawn from the statutory requirements for approval of a NADA. See 21 U.S.C. § 360b(d) (1982).
After receiving the parties’ written direct testimony and hearing cross-examination of each participant’s witnesses, the administrative law judge (AU) issued a decision finding against Cyanamid on each of the above-listed issues, and denying the supplemental application. ' Initial Decision at 94-95.
Cyanamid appealed the Initial Decision in toto to the Commissioner. Summarizing the result of his examination, the (Acting) Commissioner stated that “although Cyan-amid has [doggedly] objected to virtually all aspects of this proceeding, I am unable to conclude that any grounds exist to overturn the AU’s decision.” With some modifications, he affirmed the ALJ’s findings and conclusions. Proban Liquid: Denial of Supplemental New Animal Drug Application, No. 33-606 (June 20,1984) [hereinafter Commissioner’s Decision] at 5.
Cyanamid now petitions this court to review and reverse the Commissioner’s decision.
II. Analysis
A.
Cyanamid dominantly argues that the FDA lacks statutory authority to conduct a full safety and effectiveness review before approving a supplemental NADA. According to petitioner, the Administration must accept without renewed inspection the safety and effectiveness of an animal drug within the bounds of the previously approved NADA, and can investigate only the marginal difference between the original NADA and the supplemental NADA. In this case, therefore, Cyanamid’s interpretation would allow the FDA to consider only whether adequate directions for over-the-counter use can be written. Brief for Petitioner at 46-52, 62-71. We reject Cyanamid’s position and uphold the FDA’s thorough investigation as within its statutory authority.
The issue at hand is not novel in this court. In Rhodia, Inc. v. Food and Drug Administration, 608 F.2d 1376 (D.C.Cir. 1979), the FDA rejected Rhodia’s supplemental NADA seeking to add an alternative bulk supplier of the active ingredient in Rhodia’s product. The FDA rejection of the supplemental NADA rested on new evidence calling into question the safety of the active ingredient.
The court did not reach the issue whether the FDA, as a general matter, had the authority to pursue a full safety and effectiveness review regarding the original NADA when considering a supplemental NADA. Assuming that the FDA had such authority, the appellate panel found the administrative action in Rhodia’s case arbitrary and capricious. Id. at 1378-79. In a separate statement, however, Judge Leventhal (also the author of the majority opinion) squarely addressed the question of the Administration’s statutory authority. He considered
the FDA construction ... reasonable, and consistent with the language of § 512(e)(1)(E) of the Food, Drug and Cosmetic Act ..., which provides the primary source of FDA authority to approve supplemental NADA’s: “The supplemental application shall be treated in the same manner as the original application.” 21 U.S.C. § 360b(e)(l)(E) (1976).
The Act is to be construed liberally to effectuate its overriding purpose to protect the public health. United States v. An Article of Drug ... Bacto-Unidisk, 394 U.S. 784, 798 [89 S.Ct. 1410, 1418, 22 L.Ed.2d 726] (1969). The FDA’s broad mandate to safeguard the public health thus affords it the flexibility to shape its administrative actions when it has reason to doubt the safety of a new animal drug____
*304There are bounds to the PDA’s power to require a full safety and effectiveness review. One cannot discern the basis for such an authority where the application demonstrates that the change has no bearing whatever on issues of safety or effectiveness.
Id. at 1380-81 (footnote omitted). We find Judge Leventhal’s analysis persuasive and adopt it as our own. See also Chevron v. Natural Resources Defense Council, 467 U.S. 837, 104 S.Ct. 2778, 2782, 81 L.Ed.2d 694 (1984) (deference to agency interpretation of statute).
Cyanamid makes several attempts to avoid the sting of the FDA’s statutory authority construction. The attempts fail.
Most imaginatively, Cyanamid argues that its request for OTC status is not covered at all by the Animal Drug Amendments, but rather is governed only by 21 U.S.C. § 352(f)(1) and § 353, sections that generally concern labeling and misbranding of drugs. The argument is unconvincing. Proban’s new drug application is subject to change in accordance with the Animal Drug Amendments. See Transitional Provisions, Pub.L. 90-399, § 108(b)(2). Section 512(e)(1)(E) of the Animal Drug Amendments, 21 U.S.C. § 360b(e)(l)(E), provides that the FDA shall withdraw approval of a new animal drug application if
... the applicant has made any changes from the standpoint of safety or effectiveness beyond the variations provided for in the application unless he has supplemented the application by filing with [the FDA] adequate information respecting all such changes and unless there is in effect an approval of the supplemental application. The supplemental application shall be treated in the same manner as the original application.
We think it evident that this section encompasses Cyanamid’s supplemental application seeking OTC status for Proban. In conclusion, we add, the notion that one particular aspect of new animal drug regulation — the shift from prescription-only to OTC status — was deliberately left out of the Animal Drug Amendments runs counter to the stated purpose of the Amendments: to “consolidate the applicable provisions governing drugs ... for animal use into one logical coordinated system____” Senate Report, supra p. 2, at 3.
Cyanamid also argues that the FDA, itching to eliminate Proban entirely, is abusing review of Proban’s supplemental NADA to escape the more stringent procedures required for withdrawal of approval of a NADA. See 21 U.S.C. § 360b(e)(l)(A)-(D). Cyanamid is incorrect. The FDA’s refusal to approve OTC distribution leaves Cyanamid free to market Proban for prescription use until such time as the FDA follows the route specified for withdrawal of the original NADA. “[T]he law permits the view that concerns may exist that justify the FDA in declining to lend its approval to a strengthening of the NADA, even though the FDA is not yet warranted in terminating the NADA.” Rhodia, 608 F.2d at 1380 (additional statement of Leventhal, J.).
Cyanamid’s assertion that the FDA’s construction of its authority in this case is inconsistent with past agency policy is also without foundation. The FDA has stated and applied the full-safety-and-effectiveness-review-of-supplemental-NADAs policy on several occasions prior to the policy’s application to Proban. See Rhodia; New Animal Drug Applications: Policy on Supplemental New Animal Drug Application Approvals and Advanced Notice of Proposed Rule Making, 41 Fed.Reg. 50003 (1976) (announcing an exception to the FDA’s past policy of full safety and effectiveness review); see also In re Warner-Lambert/Parke-Davis & Co.; Benylin; Final Decision, 44 Fed.Reg. 51512 (1979) (refusing supplemental new drug application seeking OTC status for human drug after full review of safety and effectiveness data).
We conclude that it is within the FDA’s statutory authority to review a supplemental NADA affecting safety or effectiveness under the full safety and effectiveness requirements for original NADAs stated in *30521 U.S.C. § 360b(d). We now evaluate the conclusions reached by the Commissioner in his review of Proban.
B.
The burden on a NADA petitioner is not light. It includes — but is not limited to — the marshalling of: (1) adequate tests by all methods reasonably applicable to show whether the drug is safe for use under the conditions suggested in the proposed labeling; and (2) evidence consisting of adequate and well-controlled investigations, including field investigation, on the basis of which it could be fairly and reasonably concluded by experts that the drug has the effect it purports to have. See 21 U.S.C. § 360b(d)(1)(A), (E), (3). If any one of § 360b(d)’s requirements is not met, the FDA’s charge is to reject the new animal drug application. See Masti-Kure Products Co., Inc. v. Califano, 587 F.2d 1099, 1104 (D.C.Cir.1978).
The FDA found Cyanamid’s safety and effectiveness investigations on Proban inadequate in numerous respects. The ALJ and the Commissioner cited design defects in Cyanamid’s experiments and limitations on the utility of data obtained through the petitioner’s tests. Several of the shortcomings independently signaled a failure to meet a § 360b(d) requirement, and thus warranted rejection of the entire supplemental NADA.5 We need not protract this opinion by lingering over each FDA-identified deficiency in the Proban data. To demonstrate why we affirm the Administration’s disposition, it suffices to set out principal reservations regarding Cyanamid’s investigations of Proban’s safety and effectiveness indicated in the Commissioner’s decision.
Effectiveness —The Commissioner identified three major shortcomings in the efficacy data provided by Cyanamid: (1) none of the data supported Cyanamid’s claim that Proban “controlled fleas” when administered twice weekly; (2) no well-controlled field test was conducted; (3) several of the tests failed to use Proban’s market formulation. Each of these flaws supplies sufficient reason to reject the supplemental NADA.
Proban’s proposed labeling asserts that the product “controls fleas” when administered twice weekly. It was the Commissioner’s opinion that the average dogowner would take this claim to mean that Proban would provide some measure of continuous flea control for the duration of the product’s use. Commissioner’s Decision at 30-31. Cyanamid submitted data indicating that Proban killed fleas up to 24 hours after a dog ingested the drug. There was no evidence, however, of the effect Proban had on the days between applications. For all that appeared, fleas might reinfest and feed on the dog on days when no Proban was administered. Commissioner’s Decision at 135-41. The lack of evidence for Proban’s efficacy as labeled, the Commissioner determined, exterminated the possibility of approval of the supplemental NADA. 21 U.S.C. § 360b(d)(l)(E).
Cyanamid complains that the Commissioner’s reading of “control” is unreasonably stringent. We disagree. It was fully rational, we believe, to assume, as the Commissioner apparently did, that dogowners would not envision as “controlling” infestations a product working to rid a pet of fleas only two days out of seven. Continuous effectiveness between “control” dosages, we note, is hardly a novel idea. Unabated protection from dose to dose is a commonly encountered expectation when a drug or *306device is billed as one affording the user “control” over or against a condition.6
The Commission’s further ruling that Cyanamid's efficacy data must include two well-controlled studies, and that one of them must be a field study, derives from the express terms of 21 U.S.C. § 360b(d)(3) (“well-controlled investigations, including field investigation____”). The Commissioner found that Cyanamid failed to submit a well-controlled field study.
Cyanamid now asserts that the word “field” in § 360b(d)(3) is so fraught with ambiguity as to be meaningless. The petitioner’s earlier actions belie its argument. Cyanamid had labeled two studies it submitted to the FDA “field investigations.” Commissioner’s Decision at 130. Before the Commissioner, Cyanamid stated that it understood the term to mean “a study evaluating products under actual conditions of use.” Id. (quoting Cyanamid’s memorandum in support of its exceptions to the Initial Decision). Evaluation conducted in situations approximating the actual conditions of use is the definition of “field investigation” employed by the FDA, and was the definition once comprehended by Cyan-amid. We accept the FDA’s definition as a reasonable interpretation of § 360b(d)(3) and we uphold the Commissioner’s conclusion that Cyanamid did not submit the statutorily required “well-controlled ... field investigation.” See also United States v. An Article of Drug Consisting of 4,680 Pails, 725 F.2d 976, 987 (5th Cir.1984) (“[T]he very definition of substantial evidence in the Act requires that field investigations be included; an in vitro study is not a field investigation.”).
In addition to the shortfall on continuous control and the absence of a well-controlled field study, the Commissioner found several Cyanamid submissions fatally defective because they tested Proban’s active ingredient, cythioate, instead of the product’s market formulation, including its inert ingredients. The Commissioner relied on the testimony of eleven expert witnesses, including four Cyanamid witnesses, to determine that a product’s inert ingredients can affect the product’s safety and effectiveness. Commissioner’s Decision at 35-40. As the Commissioner stated, “Logic suggests that if market formulated Proban’s inert ingredients may affect product performance or safety, these, along with Pro-ban’s active ingredients, must be evaluated.” Id. at 39. We agree, and find no lack of substantial evidence for the Commissioner’s determination that “Cyanamid’s failure to use the market formulation of Proban in several of its studies renders those studies inappropriate as a basis for reaching conclusions about the properties of Proban.” Id. at 40.
Failure to use the market formulation invalidated one of the two controlled efficacy studies Cyanamid submitted for Proban. Because the FDA reasonably interpreted § 360b(d)(3) to require at least two well-controlled efficacy studies, see Commissioner’s Decision at 127, this flaw alone justifies rejection of the supplemental NADA.
Safety — In our references to the effectiveness data we have already affirmed the Commissioner three times over; we add here that similarly serious flaws mar the safety data as well. See Commissioner’s Decision at 35-38, 61-63 (failure to use market formulation); id. at 60 (rat data could not be used to determine safety in dogs); id. at 68-69 (study was biased in that, after medicated animals started consuming less food than control group, food rations for both groups were lowered); id. at 47-49 (serious questions regarding toxicity raised by test results); id. at 122-23 (risk to pregnant dogs not investigated).
In summary, upon consideration of the Commissioner's determinations and Cyan-amid’s objections to them,7 we are entirely *307satisfied that substantial evidence supports the Commissioner’s conclusion that Cyan-amid failed to present “ ‘adequate studies by all methods reasonably applicable’ demonstrating the safety of Proban for OTC use, as required by ... 21 U.S.C. § 360b(d)(l)(A)____” Commissioner’s Decision at 125.8
The FDA’s full safety and effectiveness review was within the agency’s statutory authority, and the FDA’s conclusions are supported by substantial evidence. Observing the bugs infesting Cyanamid’s safety and effectiveness data, the FDA properly rejected Proban for OTC distribution. The petition for review is accordingly
Denied.
. Known commonly as the dog flea.
. Proban was approved prior to passage of the Animal Drug Amendments of 1968. The NADA for Proban was authorized pursuant to the general section of the Food, Drug and Cosmetic Act requiring pre-marketing approval for -all new drugs, currently codified as 21 U.S.C. § 355 (1982).
. The dog flea frequently will begin its odyssey at its ultimate destination — amidst the hairs of a dog — but because the flea egg cannot fasten itself to the hairs, it soon becomes dislodged and drops to the ground (if outdoors) or to rugs, floors, upholstered furniture, etc. (if indoors). The egg hatches into a larvae form resembling a small white legless caterpillar. The larva is actually a more independent creature than the adult flea; unlike the adult, the larva is not a parasite. The prepupal flea chews on nearby dirt or debris. If the larva obtains suitable food, warmth, and moisture it will eventually *302spin around itself a cocoon of silk in which particles of debris are incorporated. Within its cocoon the larva metamorphoses into an adult flea. The adult will remain in its cocoon until vibrations or other stimuli indicate that a host is about. The wingless insect will then throw off the abode of its youth and use its powerful legs, see supra p. 1, to leap upon the potential host. H. Hoke & V. Pitt, Fleas: Their Intricate Lives and Plague-y History 36-43 (1973); Ontario Dept, of Agriculture & Food, Bed Bugs and Fleas 6 (1970).
The flea also has had an interesting literary life. See Busvine, Fleas, Fables, Folklore and Fantasies, in Fleas: Proceedings of the International Conference on Fleas 209 (R. Traub & H. Starcke ed. 1980); B. Lehane, The Compleat Flea (1969).
. "The pace of proceedings at the Food and Drug Administration ..., for whatever reasons, does not rival that of, say, a turn-of-the-century sweatshop in New York City." General Medical Co. v. FDA, 770 F.2d 214, 216 (D.C.Cir.1985) (upholding FDA’s denial of petition to reclassify perspiration inhibiting device).
. In Cyanamid's view, it is "incredible” that data the FDA considered adequate in 1968 could be regarded by the same agency as thoroughly inadequate in 1980. Brief for Petitioner at 8. We do not find the situation so implausible. As the Commissioner observed, "Scientific testing standards are capable of great change over a period of thirteen years.” Commissioner's Decision at 8. Moreover, the FDA is free to change its view if the change is supported by substantial evidence. See Contact Lens Mfrs. Assoc. v. FDA, 766 F.2d 592, (D.C.Cir.1985) (affirming FDA’s withdrawal of its own proposal to ease regulation of rigid gas permeable contact lenses).
. We would expect a consumer to be seriously disappointed, for example, to discover that a birth control pill taken at the intervals labeled did not “control” birth continuously between doses.
. In the formal argument section of its initial brief, Cyanamid does not dispute the Commissioner’s determination that Cyanamid’s safety data, by 1980 standards, was inadequate. In its statement of facts, however, Cyanamid appears to object to almost every negative finding made *307by the Commissioner. See Brief for Petitioner at 15-43. While we have treated these objections as if they had been made in the argument section, we do not recommend petitioner’s style of written advocacy.
. In his "Conclusions Regarding Proban’s Safety” the Commissioner found that Cyanamid’s supplemental NADA also failed to meet the requirements of 21 U.S.C. § 360b(d)(l)(B) and (D). We find substantial evidence supporting these conclusions.