dissenting:
I respectfully dissent.
As did the district court before it, the majority errs in failing to apply this circuit’s presumption of confusion to Silstar’s deliberate copying of the clear tip and opaque base on Shakespeare’s “Ugly Stik” fishing rods. Shakespeare maintains this color configuration as a registered trademark. Both the majority and the district court recognize that “Silstar had knowledge of, and essentially copied, Shakespeare’s trademark.” Op. at 242; Shakespeare Co. v. Silstar Corp. of America, Inc., 906 F.Supp. 997, 1010 (D.S.C.1995). With that acknowledgement, the majority should then proceed to apply the mandate of Osem Food Industries v. Sherwood Foods, Inc., 917 F.2d 161 (4th Cir.1990):
When a newcomer to the market copies a competitor’s trade dress, its intent must be to benefit from the goodwill of the competitor’s customers by getting them to believe that the new product is either the same, or originates from the same source as the *244product whose trade dress was copied. Logic requires, no less than the presumption of secondary meaning from copying, that from such intentional copying arises a presumption that the newcomer is successful and that there is a likelihood of confusion.
917 F.2d at 165 (emphasis added). It is patent that the majority simply declines to read the phrase “must be” in the above quotation from Osem Food. And it can point to no precedent in this circuit requiring a showing of something more than intentional copying before the presumption arises.* It is thus my opinion that our precedent entitles Shakespeare to a presumption of likelihood of confusion.
Had the district court properly applied this presumption, its findings of fact may well have been altered. We cannot speculate as to what those findings would have been had it employed the correct legal standard. Thus, the case should be remanded for such an examination.
With regard to Silstar’s fair-use defense, I do not believe such a defense is viable if a likelihood of confusion is established. The majority would allow a fair-use defense even when there is a likelihood of confusion, and states that it “defies logic” to hold otherwise. Op. at 243. I disagree. A leading commentator has explained the logical relationship between fair use and likelihood of confusion as follows: “The better view is that it is inconsistent to find both likely confusion and a fair use. Rather, ‘fair use’ should be viewed as merely one type of use which is not likely to cause confusion and hence is a ‘defense’ only in that sense.” 1 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 11:47 (4th ed. 1996).
The majority’s holding would place this circuit in the minority of courts that permit a fair-use defense even if a likelihood of confusion exists. See 1 McCarthy, supra, § 11:47 n. 2 (citing cases allowing fair-use defense even when likely confusion shown). I do not believe such a departure is warranted, or even permissible under circuit precedent.
In Dayton Progress Corp. v. Lane Punch Corp., 917 F.2d 836 (4th Cir.1990), we affirmed a district court which, after holding that a likelihood of confusion existed, held that “[defendant] cannot now rely on the defense of ‘fair use’ to justify its use of [plaintiffs] triliteral product designators.” Dayton Progress Corp. v. Lane Punch Corp., 12 U.S.P.Q.2d 1695, 1708 (W.D.N.C.1989). Although the majority states that the district court’s finding of bad faith, rather than its finding of likelihood of confusion, precluded the fair-use defense in Dayton Progress, that statement by the majority is hardly accurate because the Dayton Progress district court’s finding of bad faith was based on the defendant’s exact copying and identical use of the plaintiffs product. 12 U.S.P.Q.2d at 1708. On appeal, we concluded that the district court’s finding of bad faith was not clearly erroneous, and hence we accepted the district court’s factual basis for that determination. Dayton Progress, 917 F.2d at 840. Read in conjunction with the district court’s decision, as it must be, our Dayton Progress opinion supports the position that it is inconsistent to find both a likelihood of confusion and a legitimate fair-use defense when intentional copying has occurred.
Thus, the majority opinion adopts the minority view of the authorities, that a fair-use defense is available even in the face of intentional copying, and declines to follow the circuit precedent of Dayton Progress.
Finally, I should add that the penultimate rationale in the majority opinion, Op. at 243, is:
The [district] court summarized that Sils-tar sought “in good faith and not as a means of source-identification, to use a functional feature of a product” which Shakespeare had registered as a mark. Shakespeare III, 906 F.Supp. at 1016.
While the majority concludes that that factfinding is not clearly erroneous, in my opinion it is inherently incredible.
*245Accordingly, I would vacate the judgment of the district court and remand this ease for the district court to make its factual findings after applying the presumption of likelihood of confusion. I would also require the dis-triet court to correctly apply the rule with respect to a fair-use defense, as I have discussed above.
The majority refers to several cases allowing a presumption of confusion when there has been proof of intent to exploit the good will created by an already registered trademark. Op. at 240-41. These cases, however, do not directly address the effect of intentional copying and, at any rate, do not hold that intent to exploit is the only means of establishing a likelihood of confusion.