Merck & Co., Inc. (Merck) appeals from a judgment of the United States District Court for the District of Delaware (Schwartz, C.J.), 694 F.Supp. 1 (D.Del.1988), holding its U.S. Patent No. 3,882,246 (’246) unenforceable because of inequitable conduct. We affirm.

BACKGROUND

Presuming familiarity with Chief Judge Schwartz' comprehensive opinion, we describe here only the basic procedural facts and those particularly relevant to Merck’s arguments on appeal.

Merck’s ’246 patent claims a method of using cyclobenzaprine to treat certain types of skeletal muscle disorders. The ’246 patent is the result of four applications: a July 7, 1970 Canadian application; a May 21, 1971 United States application that was rejected, amended, and finally[*1419] abandoned; a January 31, 1973 continuation-in-part (c-i-p) application that was rejected, amended, and discussed in an interview at which claims 1-5 were indicated as allowable; and an April 9, 1974 c-i-p application in which those claims were again allowed. On May 6, 1975 the ’246 patent issued.

Applying in September 1970 for Food and Drug Administration (FDA) approval to sell cyclobenzaprine for use as set forth in the patent, Merck had submitted to FDA an Investigational New Drug application that contained studies of cyclobenzaprine and amitriptyline as muscle relaxants. Cy-clobenzaprine’s selectivity was emphasized. ^[1] In July 1973, Merck had submitted to FDA a report compiling, summarizing, and evaluating its cyclobenzaprine research and noting that the predominant side effect was drowsiness.

In December 1975, Merck submitted to FDA a New Drug Application (NDA) seeking approval to sell cyclobenzaprine as a skeletal muscle relaxant. Merck’s NDA submissions were consistent with the earlier submissions and referred to cycloben-zaprine’s “propensity for inducing drowsiness.” The FDA approved Merck’s sale of cyclobenzaprine as “an adjunct to rest and physical therapy for relief of muscle spasm associated with acute painful musculoskele-tal conditions.” The package insert cited drowsiness as the most frequent side effect.

On October 31, 1986, Danbury Pharmacal, Inc. (Danbury) filed an abbreviated new drug application seeking FDA approval to sell a generic version of cyclobenzap-rine for the same purpose. See 35 U.S.C. § 271(e)(2) (Supp. IV 1987). On December 17, 1986, Merck sued Danbury for infringement.

District Court Opinion

The only issues at trial were those centered on Danbury’s defenses: (1) the claimed use of cyclobenzaprine would have been obvious in view of the chemical, pharmacological and clinical properties of the prior art compound amitriptyline; and (2) the patent is unenforceable because Merck intentionally withheld prior art disclosures of amitriptyline and misrepresented cyclo-benzaprine’s side effects.

1. Obviousness

The court determined that: the Sinha publications, Lance’s work on headaches, and amitriptyline’s use in treating Parkinson’s disease all “strongly indicate amitrip-tyline’s effectiveness as a muscle relaxant”; amitriptyline and cyclobenzaprine “shar[e] a pharmacological resemblance”; “several prior art publications teach that amitriptyline and cyclobenzaprine behave similarly in a variety of tests”; and Merck’s tests indicated that amitriptyline was effective “as a skeletal muscle relaxant, but cyclobenzaprine was superior in potency and selectivity.” The court went on, however, to determine that the prior art did not disclose a reasonable expectation of cyclobenzaprine’s selectivity.

The court found that Danbury failed to show by clear and convincing evidence that the ’246 patent was invalid for obviousness, concluding:

cyclobenzaprine is distinct from previously known compounds in its selectivity. Even if amitriptyline is a selective skeletal muscle relaxant of the class of cyclo-benzaprine, no suggestion existed in the art that either would be selective in action. ... [T]he instant dispute concerns the relationship between two compounds with truly unexpected properties.... Selectivity is simply not foreshadowed in the art, much less strongly suggested.

694 F.Supp. at 32 (latter emphasis provided).

Danbury has not in this appeal challenged the determination of nonobviousness.

2. Inequitable Conduct

Dealing with the withholding of prior art, the court determined that: Merck knew of[*1420] amitriptyline’s muscle relaxant properties, and of the Sinha and Lance prior art references, but withheld all of that from the Patent and Trademark Office (PTO); Merck’s explanation for the disparity in its FDA and PTO submissions (FDA would require comparisons with amitriptyline) “strains credulity”; Merck deleted amitrip-tyline from the inventor’s 1975 published article; and Merck’s argument that ami-triptyline was immaterial was “untenable in light of the evidence.”

Dealing with the allegation of misrepresentation, the court found that Merck, in prosecuting its four applications, argued over and over that cyclobenzaprine was free of the side effects ordinarily associated with nervous system depressants; Merck’s FDA submissions and its own data indicated its awareness of at least one such side effect, i.e., drowsiness; Merck nonetheless told the PTO that cyclobenzaprine was free of “side effects of depression or causing muscle weakness and drowsiness”; and Merck’s conduct was “highly misleading, and most certainly material.”

The court held the ’246 patent unenforceable for inequitable conduct in view of FMC Corp. v. Manitowoc Co., 835 F.2d 1411, 5 U.S.P.Q.2d 1112 (Fed.Cir.1987), reasoning that:

(i) Merck withheld material prior art information concerning amitriptyline and misrepresented cyclobenzaprine’s selectivity in response to the examiner’s objections to allowing the claims; (ii) the persons responsible for the patent prosecution, Arther [Merck’s patent agent] and Share [the inventor], admittedly knew of the omitted information and of its materiality; [and] (iii) damning evidence in the form of documents and chains of events that the omissions and misrepresentation were intended to mislead the PTO.

694 F.Supp. at 35.

ISSUE

Whether the district court abused its discretion in declaring the ’246 patent unenforceable for having been inequitably procured. ^[2]

OPINION

Equitable in nature, the question of inequitable conduct is committed to the discretion of the trial court whose determination is reviewed under an abuse of discretion standard. Kingsdown Medical Consultants Ltd. v. Hollister Inc., 863 F.2d 867, 876, 9 U.S.P.Q.2d 1384, 1392 (Fed.Cir.1988) (in banc). “ ‘To overturn a discretionary ruling of a district court, the appellant must establish that the ruling is based upon clearly erroneous findings of fact or a misapplication or misinterpretation of applicable law or that the ruling evidences a clear error of judgment on the part of the district court.’ ” Id. (quoting PPG Indus. v. Celanese Polymer Specialities Co., 840 F.2d 1565, 1572, 6 U.S.P.Q.2d 1010, 1016 (Fed.Cir.1988) (Bissell, J., additional views)).

“Inequitable conduct resides in failure to disclose material information, or submission of false information, with an intent to deceive_” Kingsdown, 863 F.2d at 872, 9 U.S.P.Q.2d at 1389. Although Merck takes issue with the finding of intent, it mounts its major attack on the findings that the withheld prior art and its misrepresentations were material, saying that neither related to the selectivity of cycloben-zaprine.

1. Materiality

a. Withheld prior art

First, Merck says the determinations of nonobviousness and materiality of the withheld prior art are inconsistent, because prior art that did not render the invention claimed in the ’246 patent obvious (because selectivity was not “foreshadowed” in that art) could not be material. Second, in view of the claim limitation to[*1421] selectivity, Merck argues that the withheld prior art is merely cumulative because neither of the two references considered by the examiner were selective. See Specialty Composites v. Cabot Corp., 845 F.2d 981, 992, 6 U.S.P.Q.2d 1601, 1609 (Fed.Cir.1988); Manitowoc, 835 F.2d at 1415, 5 U.S.P.Q.2d at 1115.

First, Merck wrongly presupposes a “but for” standard of materiality. Materiality may be established, as it was here, by a showing that a reasonable examiner would consider the withheld prior art important in deciding whether to issue the patent. In re Jerabek, 789 F.2d 886, 890, 229 U.S.P.Q. 530, 533 (Fed.Cir.1986). Here, as the district court correctly found, “amitriptyline’s activity was comparable to cyclobenzap-rine’s,” 694 F.Supp. at 35, both having similar properties and effects. Thus the withheld prior art would clearly have been important to a reasonable examiner.

Second, the withheld prior art is not merely cumulative because amitriptyline was, as indicated in Merck’s own tests, by far the most relevant to skeletal muscle relaxation. To FDA, Merck disclosed ami-triptyline, not the cited prior art to which it now says amitriptyline would have been merely cumulative. That the claimed invention may have been superior in one property to both the cited and withheld prior art may be a basis for patentability; it cannot serve automatically to render the withheld prior art either cumulative or immaterial.

b. Misrepresentation

Merck says: first, its misrepresentation that cyclobenzaprine did not cause drowsiness is immaterial because drowsiness is unrelated to selectivity; second, its single statement on drowsiness has been taken out of context because drowsiness in proper context refers to drowsiness accompanying muscle weakness; third, absence of a drowsiness effect was not relied on by the examiner in deciding to allow the ’246 patent; and fourth, it did not misrepresent cyclobenzaprine’s unique selectivity.

First, the court found that cyclobenzap-rine’s selectivity was related to the side effect of drowsiness. It stated that “[t]o the extent drowsiness falls within the spectrum of sedation, it appears ... that cyclo-benzaprine produces sedative effects, although not muscle weakness,” and thus “cyclobenzaprine’s selectivity is limited to its operation without loss of normal muscle tone.” Merck has not shown that finding to have been clearly erroneous or that the district court abused its discretion in so viewing the evidence.

Second, the district court did address Merck’s single June 1972 description of cy-clobenzaprine’s action as free of “the attendant side effects of depression or causing muscle weakness and drowsiness.” The court’s reference to that one statement as the most glaring misrepresentation, however, was not out of context and does not limit its misrepresentation findings to that one statement. The entirety of the district court’s opinion recognizes the misrepresentations of cyclobenzaprine’s side effects made throughout the patent prosecution: “[ejqually disturbing is Merck’s blatant misrepresentation of the side effects of cyclobenzaprine in response to the examiner’s reluctance (to put it mildly) to allow the claims.” 694 F.Supp. at 35. Merck was aware of the side effect of drowsiness and that it clearly resembles side effects ordinarily associated with nervous system depressants. Thus the court correctly cited a pattern of misrepresentations regarding cyclobenzaprine’s freedom from the attendant side effects ordinarily associated with nervous system depressants.

Third, Merck again improperly argues for a “but for” standard of materiality in its assumption that the examiner did not rely on its drowsiness misrepresentation. To be material, a misrepresentation need not be relied on by the examiner in deciding to allow the patent. The matter misrepresented need only be within a reasonable examiner’s realm of consideration.

Fourth, in view of cyclobenzaprine’s and amitriptyline’s comparable activity reflected in Merck’s tests, we cannot say the court abused its discretion in finding that Merck’s representation on the invention’s[*1422] unique selectivity “was not entirely accurate”. Whether that representation must be deemed a fatal misrepresentation is of little or no moment, however, in light of Merck’s pattern of misrepresentations.

2. Intent

Intent need not, and rarely can, be proven by direct evidence. It is most often proven by “a showing of acts the natural consequences of which are presumably intended by the actor.” Kansas Jack, Inc. v. Kuhn, 719 F.2d 1144, 1151, 219 U.S.P.Q. 857, 861 (Fed.Cir.1983). In light of all the evidence rightly described by the court as “damning”, including Share’s and Arther’s knowledge of the highly material amitripty-line prior art, the simultaneous submission of amitriptyline data to FDA and its withholding from the PTO, the material misrepresentation to the PTO of cyclobenzaprine’s side effects, and the decision to delete the amitriptyline portion of the inventor’s 1975 article, we cannot say that the court’s finding of intent to mislead was clearly erroneous. Nor can we conclude that the court abused its discretion in determining that Merck was guilty of inequitable conduct.

AFFIRMED.