FRIEDMAN, Circuit Judge.
This case involves electric shavers. Windmere Corporation (Windmere) appeals from a final order of the United States District Court for the Southern District of Florida directing a verdict in favor of U.S. Philips Corp. and North American Philips (collectively Philips) on Windmere’s antitrust counterclaims, 680 F.Supp. 361. Because we hold that there was sufficient evidence to submit the antitrust claims to the jury, we reverse and remand for a new trial on those claims.
[697]*697I
A. The Proceedings in the District Court.
In October 1984, Philips filed suit in the district court against Windmere and Izumi Seimitsu Kogyo Kabushiki Kaisha (hereinafter Izumi) (a Japanese corporation) for infringement of Philips’ electric shaver patent and against Windmere for unfair competition. Windmere filed a counterclaim alleging that Philips had monopolized and attempted to monopolize the electric rotary shaver market, in violation of Section 2 of the Sherman Act, 15 U.S.C. § 2 (1982), and that Philips and Dutch Philips (N.V. Philips Gloeilampenfabrieken) had conspired to restrain trade in that market, in violation of Section 1 of that Act, 15 U.S.C. § 1 (1982). Windmere apparently has abandoned its Section 1 claim.
Both the antitrust and the patent issues were tried to a jury. At the close of Wind-mere’s antitrust case, the district court granted Philips’ motion for a directed verdict on those issues. The court submitted the patent and unfair competition issues to the jury, which ruled for Philips on the patent claim and for Windmere on the unfair competition claim. Pursuant to Rule 54(b) of the Federal Rules of Civil Procedure, the court certified its order directing a verdict on the antitrust counterclaim as final and therefore appealable.
B. The Evidence Relating to the Antitrust Counterclaim.
1. The product involved. Philips sells, among other things, electric shavers under the NORELCO and the SCHICK brand names. There are two styles of electric shavers: rotary and foil. Rotary shavers use high-speed rotating cutters under a circular comb. Foil shavers use vibrating blades under an etched metal foil. Rotary shavers are either cordless rechargeable models or less expensive corded models.
Philips introduced its Norelco rotary shaver in 1948. The first model contained a single head. This was followed by a double- and then a triple-headed shaver; Philips introduced the latter in 1966. The Norelco triple-headed rotary shaver accounts for at least 75 to 80 percent of Philips’ total shaver sales. In 1981, Philips purchased the rights to the SCHICK trademark and used it to market its foil shaver.
2. The rotary shaver market.
a. The relevant market. Philips does not dispute that there was sufficient evidence upon which the jury could have concluded that for antitrust purposes electric rotary shavers constituted a separate market from electric foil shavers. There also was evidence that shavers priced above $30 constituted a separate market from those priced below that amount. Philips has tended to concentrate on the higher priced models, which generally are the cordless or rechargeable shavers. In 1983, the year before Windmere entered the electric shaver market with its Ronson shaver, Philips had 72.8 percent of the sales of above-$30 shavers.
b. The concentration of that market. Prior to the entry of Windmere and Sears Roebuck into the rotary shaver market, Philips’ Norelco shaver was the only product there. Mr. Gould, Windmere’s economist expert, who had worked for the Census Bureau and the Department of Justice, estimated that “in 1985 the Norelco share of the rotary market was of [sic] the order of ninety percent.” Gould, who had studied the concentration ratios of more than 400 industries, testified that Norelco’s share of the rotary market was “the highest figure I have encountered in all of my studies_ Going back thirty years.”
c. Barriers to entry. There was evidence that a recognized brand name was critical to successful entry into the electric shaver market. In 1977, Windmere first attempted to market a foil shaver and used the manufacturer’s name (Payer). The name was not well-known, and the shaver line was discontinued. The president of Windmere testified that Windmere’s customers advised them “that there was no way we could be in the electric shaver industry without a brand. Windmere doesn’t qualify as a national brand that was known in the electric shaver business which is what they told us what [sic] we [698]*698needed.” Although this testimony was admitted with a limiting instruction that it was allowed to show only “state of mind” and not the “truth of the answer,” the fact that Windmere believed that it needed a brand name to succeed in the market is evidence that such a name was necessary.
Philips became aware in February 1980 that Windmere had reached “an understanding” to acquire the SCHICK brand name. Philips subsequently predicted that if anyone bought the SCHICK name and used it to market Izumi rotary shavers, as Windmere was attempting to do, Philips would lose $15 million in sales and $2 million in profit during a three-year period, as well as five points in market share. Philips then bought the rights to the SCHICK brand name and used it to begin marketing a line of foil shavers. A Philips memorandum titled “FOIL SHAVER MARKETING OPPORTUNITY” noted that one “of the factors which have influenced our decision not to market this line [foil shavers] in the past” was because
we [Philips] have been unwilling to commit the tens of millions of dollars required to establish an entirely new shaver brand in the U.S. market.
Now, however, we are faced with a unique opportunity which may never come our way again. An established brand name is available for a relatively small investment. Virtually all roadblocks which have stood in the way of the foil line are eliminated by the availability of the brand name — SCHICK.
Philips’ own economist expert testified that entry into Windmere’s proposed markets was relatively easy. Philips also relied on the fact that following Windmere’s entry into the rotary shaver market in 1984, Windmere obtained 3 to 5 percent of the total electric shaver market.
d. Windmere’s entry into the rotary shaver market and Philips’ response. As noted, in 1977 Windmere unsuccessfully had attempted to enter the foil shaver market. In the spring of 1984, Windmere again entered the electric shaver market with both foil and rotary shavers manufactured in Japan, which Windmere sold under the RONSON trademark.
Windmere introduced a cordless rotary (model RR-1) shaver for $34.99 and a corded rotary (model RR-2) shaver for $22. Windmere projected that, during its first six months in the shaver business, it would sell 125,000 razors for total dollar sales of more than $3 million. In the last nine months of 1984, Windmere actually sold more than 300,000 units for approximately $6 million. Windmere’s 1984 annual report, prepared after it had filed its antitrust counterclaim, stated that “[d]espite a substantial loss of sales and revenues resulting from Philips, i.e., Norelco’s business tactics, the Ronson electric shavers represent the most successful new product line introduction in the history of the Company.”
Philips had described Windmere as “a financially extremely healthy enterprise which is capable of making life very difficult for us.” A Philips “report of decision taken at a management policy meeting” on January 5, 1984, stated: “If Windmere launches rotary shavers on the U.S. market, Mr. Beasley [vice president of marketing for North American Philips Consumer Product Division] proposes to kill this stone dead by introducing old models at very low prices.... ”
After Windmere introduced its Ronson shavers at a price lower than Philips’ shavers, an April 20,1984 Philips weekly report, for the week ending April 21, 1984, stated: “Ronson is getting a solid foothold and apparently is willing and able to pay a high price to get into our business. Let’s pound them into the sand.” The report suggested that Philips introduce the model “1320” cordless rotary shaver. The model 1320 was Philips’ old model 1312, which had an expected retail price of $44 during 1979 to 1980.
The model 1320, however, was introduced at a distributor price of $27.08, which was approximately 31 percent less than the $39 price at which the same razor had been sold as the model 1312 four years earlier, and approximately 23 percent less than the comparable Ronson model. The [699]*699introduction of the model 1320 was not mentioned in Philips’ 1984 business plan, dated December 1983. Until the 1320, Philips had never introduced a new product at a lower price point.
Philips discussed the 1320 shaver in a March 1984 memorandum:
Until later in the year, Norelco has decided on an absolute embargo on any form of publicity in connection with the introduction of the HP 1320.... The negative effect of any premature release of information on the HP 1320 or 1337 is feared, just at the very time when the intention is to introduce the 1320 as a complete addition to the line and not for it to take sales away from any of the existing range of standard shavers.... The question was asked whether it was the intention to include the 1320 in the line for 1985. Beasley and Co. have taken no clear position on this subject. Given the likely impossibility of driving Windmere off the U.S. market completely, we foresee.that a non-retracting rechargeable shaver sold at an aggressive price level will persist on the market for a rather longer time and thus cannot be viewed as an incidental phenomenon, so to speak (emphasis added).
Philips later described the 1320 model as a “200th anniversary special to be out for a limited period of time.”
A May 24, 1984 Philips memorandum stated: “As discussed at the policy meeting, we should attempt to formulate a short term strategy with regard to the competition represented by the Ronson trademark, which in the hands of Windmere Inc. of Florida has penetrated our market segment.” The memorandum refers to a marketing strategy wherein
[ a] direct frontal attack on Ronson in the marketplace is probably the crudest and probably also one of the most expensive means at our disposal. This would involve, for example, us locating competing dealers at every address where Ronson is sold and selling Norelco and/or Philips shavers at a lower price. Furthermore, every challenge issued by Ronson in the form of actual discounts would have to be countered by offering even lower discounts.
We can reward exclusivity by giving every dealer who does not sell Ronson shavers an extra discount of 5%.
Total attack strategy, that is to say that just as Norelco is now doing, we would launch a multiplicity of activities against Windmere and at a later stage against any other competitor who appears on the scene.
The memo concludes that “legal measures” “should make life difficult for Ronson on every front where it is at all possible, on the basis of our patents” and that as a “tactical measure” Philips should “[b]uy Izumi, and quickly.”
Windmere presented testimony that, after Philips introduced its model 1320, substantial orders ($2-2.5 million) were can-celled. No orders for Ronson products had been cancelled before the Philips 1320 was introduced.
The 1984 Philips marketing plan included a model 1615 corded rotary shaver. The 1615 was to have been sold to distributors for $27.99, who then would have sold the shavers at a retail price of $29.99. The price of the 1615, however, was dropped approximately 35 percent to a distributor price of $18.03, which was almost $4 less than the distributor price for Ronson’s comparable corded rotary shaver.
Philips’ response to Windmere’s entry affected other models in its line. In its “Model policy 1985,” discussed in May 1984, a twin-head model “1722” apparently was scrapped:
Mainly on account of the action undertaken against Ronson as well as because of the very presence of Ronson on the market with its known price advantages, there is less room for this product on the American market than had been originally planned.
Another model, the “1622,” was also affected:
Although ... this should be an expensive model on account of the enhanced specification compared with the HP 1620, it [700]*700appeared from the discussion of this point that the matter of price was more important than the question of the specification. Norelco maintained their view despite the poor way in which sales of the HP 1620 had held up, even in the USA as well. Later the product discussion turned to the question of a frontal attack on Ronson. In this context, Gron-ingen [Dutch Philips] quite clearly declared it was ready to drop the price of this product even below that of the HP 1620 ($8.15) provided that Norelco were ready to drop their gross margin by a considerable amount. Such a frontal attack on Ronson must certainly be accompanied by considerably more volume....
In March 1985, one year after Windmere entered the electric shaver market, a Philips memo of a sales meeting and visits to buyers noted the damaging effect the 1615 and 1320 had on sales of other Philips models:
The HP 1615 and the HP 1320 clearly eat in on the HP 1337, HP 1327D and the HP 1604/05/06. This is a dramatic development since it appears that the HP 1615, at a price point of $19.95 will be a disturbing development. It is clear that this is a planned development and that we cannot occupy ourselves “crying over spilled milk,” but the fact remains nonetheless.
e. Philips’ alleged predatory pricing. Windmere contended that Philips sold its 1320 and 1615 models below its average variable costs and sacrificed its short-term profits on those models in order to drive Windmere out of the rotary shaver market. Windmere and Philips introduced into evidence various calculations directed to the question whether Philips sold below cost. As the parties framed the issue, the answer depended upon which of Philips’ costs properly were allocable to those two models.
The parties vigorously disputed whether Philips’ advertising expenses were a variable cost of those models and, if they were, the basis upon which advertising expenses should be allocated. Philips contended that there should be subtracted from the $17.5 million it spent for advertising in 1985 the approximately $6.6 million it received for advertising in that year from its Dutch parent, N.V. Philips. If that subtraction were made, there was evidence that Philips made a profit on the sales of the shavers. Philips also argued that advertising expenses should have been allocated as a percentage of expenses based upon revenues — an allocation under which Philips would have made a profit. Windmere disputed these assumptions and calculations based on them. Each side presented evidence justifying and supporting its position.
The parties also disputed the allocation of costs other than advertising. For example, Windmere relied upon its projections of Philips’ 1985 costs of batteries, freight and handling. Philips contended that based on its actual, as distinguished from projected, costs for those items, it realized a profit on its sales. Windmere responded that because Philips’ actual costs were not available to it, its calculations based upon projected costs were a reasonable basis for determining whether Philips’ sales were at a profit.
f. Windmere’s withdrawal from the rotary shaver market. In December 1985, less than two years after it introduced its Ronson rotary shaver, Windmere announced its withdrawal from the rotary shaver market. (It remained in the foil market.) Shortly thereafter, Philips discontinued the sale of its 1320 and 1615 rotary shavers. At a 1986 housewares show, Philips stated that “[t]hey were closing [the model 1320] out and coming out with a new model at a higher price.”
C. The Decision of the District Court Directing a Verdict on the Antitrust Counterclaim.
In an opinion explaining “in detail” its “reasons for granting” Philips’ motion for a directed verdict, the district court stated that “reluctantly” it had found that “Wind-mere had failed to present substantial evidence sufficient to create a jury question on at least one element of its antitrust claims — the willful acquisition or mainte[701]*701nance of monopoly power via predatory pricing_” (footnote omitted).
The court ruled that “Windmere presented a substantial amount of evidence to support its definition of the relevant market and to show that [Philips] had market power in this market.” The court stated that “[t]he specific conduct which Windmere alleged was anticompetitive and showed Philips’ intent to monopolize, was Philips[’] alleged predatory pricing scheme with regard to the sale of its model 1320 and 1615 rotary electric razors.” The court held that under International Air Industries, Inc. v. American Excelsior Co., 517 F.2d 714 (5th Cir.1975), cert. denied, 424 U.S. 943, 96 S.Ct. 1411, 47 L.Ed.2d 349 (1976), Windmere was required to show that Philips was “ ‘charging a price below [its] average variable cost in the competitive market,’ ” (quoting 517 F.2d at 724) and that “Windmere’s evidence of below-cost pricing does not meet the substantial evidence standard necessary to withstand a directed verdict.”
The court stated:
The disputes regarding the evidence Windmere introduced to prove predatory pricing involve two primary issues: (1) the proper allocation of Philips[’] advertising expenditures, and (2) the proper method of proving costs — i.e., by using actual costs or projections.
The court further stated that it was “undisputed that if all of Philips[’] advertising expenditures are classified as fixed costs, that Philips[’] prices for the model 1320 and 1615 razors exceeded their average variable cost.” The court noted the testimony of Mr. Kamerschen, Philips’ expert, that “based on his review of Norelco’s business and media plans, which he stated contained the type of information an economist would generally rely upon in order to make a judgment of this kind, that Norelco’s advertising was a fixed cost,” and stated that “Windmere failed to bring forth any information to the contrary on the cross-examination of this witness” (record reference omitted). The court ruled that the evidence upon which Windmere relied to show that advertising was a variable expense — Norel-co’s “1984 income statements ... which listed advertising in a category entitled variable expenses” — was insufficient because “such an accounting record cannot alone support the classification of advertising as a variable expense for economic purposes” (record reference omitted).
The court further held that “[e]ven assuming, however, that advertising expenses here are properly characterized as variable costs, there is a question whether cooperative advertising support given to Norelco by NVP should be deducted before determining whether Norelco priced the 1615 and 1320 below its average variable cost.” It rejected, as “wholly unsupported” by the evidence, Windmere’s “contention” that the approximately $6.6 million Philips received “in advertising support from NVP in 1985” either was not actually paid or was in addition to and not a part of the $17.5 million Norelco spent for advertising in 1985. The court also held that Windmere had not introduced sufficient evidence to justify submitting to the jury Windmere’s contention that the advertising expenses should be allocated to the 1615 and 1320 models “on a per unit basis” (footnote omitted). Finally, the court rejected Windmere’s evidence of the cost of batteries, freight and handling because that evidence was based on “projections of the 1985 year” (emphasis in original).
II
A. This court has exclusive jurisdiction over an appeal from a final judgment where the jurisdiction of the district court was based in whole or in part on 28 U.S.C. § 1338(a) (1982). 28 U.S.C. § 1295(a)(1) (1982). In filing its complaint against Windmere and Izumi for patent infringement, Philips invoked the jurisdiction of the district court under section 1338(a). Although Windmere’s antitrust counterclaim, which is the subject of this appeal, did not invoke the district court’s jurisdiction under that section, our appellate jurisdiction depends upon the nature of the case in the district court and not upon the issues presented to us for review. See In re Innotron Diagnostics, 800 F.2d 1077, 1080, [702]*702231 USPQ 178, 180-81 (Fed.Cir.1986) (“the mere presence in a case of ‘issues’ other than those within a substantive field within this court’s exclusive appellate jurisdiction [does not] oust this court of the jurisdiction it must and would otherwise exercise in carrying out its mission.”); see also Atari, Inc. v. JS & A Group, Inc., 747 F.2d 1422, 1431, 223 USPQ 1074, 1080 (Fed.Cir.1984) (en banc).
In the present case, the jurisdiction of the district court over this case was based in part on section 1338(a). That fact gives us jurisdiction over this appeal, even though the appeal presents no questions under the patent laws. See, e.g., In re Innotron Diagnostics, 800 F.2d at 1080, 231 USPQ at 180 (where patent complaint was consolidated with antitrust complaint, the Federal Circuit had exclusive jurisdiction over appeal from final judgment); Cable Elec. Prods., Inc. v. Genmark, Inc., 770 F.2d 1015, 1029, 226 USPQ 881, 890 (Fed.Cir.1985) (where district court exercised jurisdiction over patent count under § 1338(a), the Federal Circuit has jurisdiction, under § 1295(a), to review the appeal of the nonpatent counts).
B. In determining whether the district court correctly directed a verdict on Wind-mere’s antitrust counterclaim, we apply the law of the regional circuit in which the district court sits to determine both the appropriateness of a directed verdict and the antitrust law questions. Power Controls Corp. v. Hybrinetics, Inc., 806 F.2d 234, 237, 231 USPQ 774, 776 (Fed.Cir.1986); Loctite Corp. v. Ultraseal Ltd., 781 F.2d 861, 875, 228 USPQ 90, 99 (Fed.Cir.1985) (“We must approach a federal antitrust claim as would a court of appeals in the circuit of the district court whose judgment we review.”); Panduit Corp. v. All States Plastic Mfg. Co., 744 F.2d 1564, 1574-75, 223 USPQ 465, 471 (Fed.Cir.1984).
We thus “apply [Eleventh] Circuit law primarily and look to guidance (albeit nonbinding) from other circuits, especially in uncharted areas.” 781 F.2d at 875, 228 USPQ at 99-100. Since the Eleventh Circuit treats as binding decisions of the Fifth Circuit rendered before October 1, 1981, when the Eleventh Circuit came into being, Bonner v. City of Prichard, Ala., 661 F.2d 1206, 1207 (11th Cir.1981) (en banc), we also apply those earlier Fifth Circuit decisions.
In the Eleventh Circuit, “[t]o avoid a directed verdict, [Windmere] must have presented sufficient evidence to create a jury question with respect to each element of [its] case.” Malcolm v. Marathon Oil Co., 642 F.2d 845, 848 (5th Cir. Unit B April), cert. denied, 454 U.S. 1125, 102 S.Ct. 975, 71 L.Ed.2d 113 (1981).
In determining whether there is sufficient evidence to submit a ease to the jury following a motion for directed verdict, the district court should consider all evidence and not just evidence supporting the non-mover’s case. The evidence must be considered in the light and with all reasonable inferences most favorable to the party opposed to the motion. If the facts and inferences point so strongly and overwhelmingly in favor of one party that reasonable persons could not reach a different conclusion, the Court should direct a verdict.
Gregory v. Massachusetts Mut. Life Ins. Co., 764 F.2d 1437, 1440 (11th Cir.1985). See also Boeing Co. v. Shipman, 411 F.2d 365, 374 (5th Cir.1969) (en banc) (for evidence to be sufficient to create a jury question it must be of a quality and weight that reasonable, impartial, and fair-minded jurors might reach different conclusions).
In Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986), the Supreme Court held that the standard for granting summary judgment under Rule 56(c) of the Federal Rules of Civil Procedure “mirrors the standard for a directed verdict under Federal Rule of Civil Procedure 50(a), which is that the trial judge must direct a verdict if, under the governing law, there can be but one reasonable conclusion as to the verdict. If reasonable minds could differ as to the import of the evidence, however, a verdict should not be directed.” Id. at 250-51, 106 S.Ct. at 2511 (citation omitted). The Court pointed out that “[credibility determinations, the weighing of the evidence, and the [703]*703drawing of legitimate inferences from the facts are jury functions, not those of a judge, whether he is ruling on a motion for summary judgment or for a directed verdict. The evidence of the non-movant is to be believed, and all justifiable inferences are to be drawn in his favor.” Id. at 255, 106 S.Ct. at 2513.
Ill
“The offense of monopoly under § 2 of the Sherman Act has two elements: (1) the possession of monopoly power in the relevant market and (2) the willful acquisition or maintenance of that power as distinguished from growth or development as a consequence of a superior product, business acumen, or historic accident.” United States v. Grinnell Corp., 384 U.S. 563, 570-71, 86 S.Ct. 1698, 1703-04, 16 L.Ed.2d 778 (1966). The district court recognized, and Philips does not dispute, that “Wind-mere presented a substantial amount of evidence to support its definition of the relevant market [rotary electric shavers] and to show that [Philips] had market power in this market.” The question before us, therefore, is whether Windmere presented sufficient evidence to entitle it to go to the jury on the second element of monopolization: whether Philips willfully had maintained its monopoly power in the rotary shaver market.
The district court viewed the essence of Windmere’s case on the second element as “the willful acquisition or maintenance of monopoly power via predatory pricing,” and used, as the standard for determining “predatory pricing,” pricing “below average variable cost.” The court concluded that Windmere had failed to present substantial evidence that Philips had priced its 1615 and 1320 rotary shavers below Philips’ average variable costs.
We conclude, however, that the district court took too narrow a view of the kind of evidence that will support the second element of monopolization and of the bases of Windmere’s claims. Moreover, even under the district court’s narrow view of Windmere’s theory, Windmere presented sufficient evidence of predatory pricing to preclude a directed verdict on that issue.
A. There was evidence from which the jury could have concluded:
1. That entry barriers to the rotary electric shaver market are substantial, if not high — the need to have a well-known brand with wide consumer acceptance, the limited number of brands that satisfy this requirement, and the substantial advertising expenditures required to attain a foothold in the market.
2. That Philips introduced its lower-priced older models only in response to Windmere’s entry into the market with its cheaper Ronson rotary electric shaver.
3. That Philips cut the price of its shavers in an endeavor to eliminate Windmere from the market. Mr. Beasley proposed that if Windmere started selling rotary shavers in the United States market, Philips should “kill this stone dead by introducing old models at very low prices....” A Philips weekly report, made after Philips had introduced its rotary shaver, suggested: “Let’s pound them into the sand.” A Philips memorandum referred to “the likely impossibility of driving Windmere off the U.S. market completely_” Although Philips attempts to discredit this evidence as mere sales talk, it was for the jury to determine what weight and significance this evidence had.
That this purpose of Philips was further shown by the evidence that, when Wind-mere withdrew from the rotary shaver market, Philips promptly discontinued the sale of its lower-priced old model.
4. That Philips’ willful maintenance of its monopoly position was further shown by Philips’ prior acquisition of the SCHICK brand name after it had learned about an understanding between Windmere and Schick for Windmere to acquire that name.
To be sure, Philips introduced evidence designed to show that its actions were legitimate competitive responses prompted by the entry of Windmere into the market. The evaluation of all the evidence bearing on the nature of Philips’ actions and the inferences to be drawn from the evidence, [704]*704was for the jury to determine. If the jury had credited Windmere’s evidence and rejected Philips’, it could have returned a verdict for Windmere. Evidence that a firm holding 90 percent of a market that has substantial entry barriers drastically slashes its prices in response to the competition of a new entrant, for the purpose and with the effect of eliminating that entrant, is sufficient to show monopolization, in violation of section 2 of the Sherman Act. Cf. Aspen Skiing Co. v. Aspen Highlands Skiing Corp., 472 U.S. 585, 610-11, 105 S.Ct. 2847, 2861-62, 86 L.Ed.2d 467 (1985); American Tobacco Co. v. United States, 328 U.S. 781, 66 S.Ct. 1125, 90 L.Ed. 1575 (1946).
B. Even under the district court’s view that Windmere’s case rested on its claim that Philips had sold below average variable cost, the evidence was sufficiently in conflict to require submission of the issue to the jury and precluded a directed verdict. As the district court noted, the determination whether Philips sold below average variable cost required resolution of two issues: (1) whether Philips’ advertising expenses were variable or fixed costs and, if variable, whether the advertising allowance Philips received in 1985 from its Dutch parent should be deducted from Philips’ own advertising expenditures; and (2) whether Windmere’s projections for certain items of Philips’ cost — batteries, freight and handling — were sufficient to establish those costs. The district court held that on both of these elements of its case, Wind-mere failed to present sufficient evidence to create disputed factual issues for the jury. We disagree.
1. In ruling that Windmere had not presented substantial evidence that Philips’ advertising expenses were variable rather than fixed costs to create a jury question on that issue, the district court relied on the testimony of Philips’ economist expert, Mr. Kamerschen, that the Norelco advertising was a fixed cost, and stated that “Windmere failed to bring forth any information to the contrary on the cross-examination of this witness.” It was for the jury, however, and not for the court, to determine the credibility of this witness and the weight to be given his testimony. The jury was not required to accept his expert testimony, even if it was uncontradicted. See Quock Ting v. United States, 140 U.S. 417, 420, 11 S.Ct. 733, 734, 35 L.Ed. 501 (1891) (“There may be such an inherent improbability in the statements of a witness as to induce the court or jury to disregard his evidence, even in the absence of any direct conflicting testimony.”). See also Archer v. Commissioner, 227 F.2d 270, 273 (5th Cir.1955); Boyett v. Commissioner, 204 F.2d 205, 208 (5th Cir.1953).
To counter Mr. Kamersehen’s opinion, Windmere relied on Norelco’s 1984 income statements, “which listed advertising in a category entitled variable expenses.” The court rejected this evidence as insufficient to create an issue for the jury because “[i]t is clear that such an accounting record cannot alone support the classification of advertising as a variable expense for economic purposes.” As was the case with Mr. Kamerschen’s expert testimony, however, this was evidence that the jury was entitled to consider in determining whether the advertising expenses were variable or fixed costs. Indeed, the Areeda and Turner antitrust treatise, which the district court cited as authority for rejecting Philips’ own characterization of its advertising expenses as variable costs, states that “we have defined marginal and average variable cost to include advertising and other promotional expenditures.” 3 P. Areeda & D. Turner, Antitrust Law 11721a, at 191 (1978).
Similarly, we conclude that, in determining the amount of Philips’ advertising expenses that are variable costs, the proper treatment of the approximately $6.6 million Philips received from its Dutch parent in 1985 was an issue for the jury to decide.
2. The question whether Philips’ advertising expenses should be allocated to the 1615 and 1320 models on a per unit basis or on the basis of revenue received also was a matter for the jury and not for the court to decide. We reach that same conclusion with respect to the propriety of Windmere’s use of 1985 projections, rather [705]*705than Philips’ actual operating costs (which apparently were not available to Wind-mere), to determine Philips’ costs of batteries, freight and handling.
IY
It is unfortunate that, as a result of our decision, the lengthy and complex issues involved in Windmere’s antitrust counterclaim will have to be retried. That, however, is the consequence of the district court’s directing a verdict rather than allowing the case to go to the jury. As the Eighth Circuit stated, in language that applies equally to this case:
In short, the case was not clear enough to justify a directed verdict. If this case had been tried to the court, we would have had no difficulty in affirming, as not clearly erroneous, a finding for [Philips] supported by the opinion the District Court rendered here. But this case was not tried to the court. It should have been permitted to go to the trier of fact.
Dace v. ACF Indus., Inc., 722 F.2d 374, 379 (8th Cir.1983). The court continued in a footnote:
This case illustrates again that it is usually better practice for a district court, faced with a motion for directed verdict, to allow the case to go to the jury, and address the issue by way of judgment n.o.v. if necessary. The jury may find for the moving party, in which case the issue disappears. If the verdict is against the moving party, and if judgment n.o.v. is granted, and if this Court decides on appeal that it should have been denied, the verdict can simply be reinstated, and no new trial is necessary.
Id. at 379-80 n. 9. Accord Mattivi v. South African Marine Corp., “Huguenot”, 618 F.2d 163, 166 n. 2 (2d Cir.1980); Jamesbury Corp. v. Litton Indus. Prods., Inc., 839 F.2d 1544, 1550, 5 USPQ2d 1779, 1784 (Fed.Cir.1988).
CONCLUSION
The order of the district court granting a directed verdict against Windmere on its antitrust counterclaim is reversed, and the case is remanded to the district court for a new trial on that counterclaim.
REVERSED AND REMANDED.