We must decide whether a former collegiate soccer player may pursue her federal antitrust challenge to an intercollegiate athletic association rule that discourages student-athletes from transferring to member institutions during the course of their collegiate athletic careers.

I

A star high school soccer player, Rhiannon Tanaka (“Tanaka”) was heavily recruited by the athletic programs of a number of universities, including the University of Southern California (“USC”), which belongs to the Pacific-10 Conference (“Pac-10”). The Pac-10 is an association of ten universities which was formed for the purpose of “establishing an athletic program to be participated in by the members.” Hairston v. Pacific 10 Conference, 101 F.3d 1315, 1317 (9th Cir.1996). During her senior year of high school, Tanaka met with USC athletic officials, including Michael Garrett and Daryl Gross, who allegedly made certain representations to her regarding the USC women’s soccer program. She also inquired about transfer restrictions, and was allegedly told that she would be free to transfer without penalty provided she remained at USC for one year and met minimum academic requirements. Tana-ka thereupon signed a letter of intent to enroll at USC and attended USC for the 1994-95 academic year.

Tanaka quickly became dissatisfied with the state of USC’s women’s soccer program and the quality of her USC education. In particular, she claims that USC was arranging for athletes to receive fraudulent academic credit through sham classes. In the spring of 1995, she received permission from USC to communicate with other schools about transferring to their programs. Because she wished to remain in Los Angeles, and because its women’s soccer program was nationally ranked, she decided to transfer to the University of California, Los Angeles (“UCLA”), another Pac-10 member institution.

USC opposed Tanaka’s transfer to UCLA, however, and sought sanctions against her pursuant to Pac-10 Rule C 8-3-b (“transfer rule”), which governs intra-conference transfer. The rule provides, in pertinent part:

USC insisted that Tanaka sit out her first year at UCLA and lose one year of athletic eligibility. Tanaka unsuccessfully appealed her sanction. In addition to her loss of athletic eligibility, Tanaka was denied any athletically related financial aid during her first semester at UCLA.

Tanaka alleges that USC invoked the transfer rule sanctions against her in retaliation for her participation in an investigation into possible academic fraud involving student-athletes at the school. In fact, her complaint expressly alleges that she “is the only transferring athlete who had the sanctions imposed on her” (emphasis in original), and further that “[i]n all other instances where student-athletes transferred from USC, Pacific-10 penalties had not been used.”

In May 1998, Tanaka filed an unsuccessful state-court action for fraud against USC. On May 5, 1999, Tanaka filed the instant action, asserting a state breach of contract claim and a claim under the Clayton Act, 15 U.S.C. § 15 , predicated on a violation of section one of the Sherman Act, 15 U.S.C. § 1 . The district court dismissed the complaint in August 1999 with leave to amend. Tanaka filed her first amended complaint on September 13,1999. On November 15, 1999, the district court again dismissed, this time with prejudice. The court held that the Pac-10 transfer rule was beyond the reach of the Sherman Act because it was essentially “noncommercial” in nature, reasoning the rule “is more tied to defendants’ noncommercial rather than commercial activities.” The court noted that, even if the Sherman Act applied, Tanaka would likely still lose, as the transfer rule would not be found unreasonable under the rule of reason. The court declined to exercise supplemental jurisdiction over Tanaka’s state contract claim pursuant to 28 U.S.C. § 1367 (c)(3). This timely appeal followed.

II

A

Section one of the Sherman Act (“Act”) provides in pertinent part:

15 U.S.C. § 1 . In order to establish a claim under Section 1, a plaintiff must demonstrate: “ ‘(1) that there was a contract, combination, or conspiracy; (2) that the agreement unreasonably restrained trade under either a per se rule of illegality or a rule of reason analysis; and (3) that the restraint affected interstate commerce.’ ” Hairston, 101 F.3d at 1318 (9th Cir.1996) (quoting Bhan v. NME Hosp., Inc., 929 F.2d 1404 , 1410 (9th Cir.1991)). The district court concluded that Tanaka failed to meet the second part of this test, reasoning that the transfer rule is essentially “noncommercial,” and hence does not involve “trade.” Thus, according to the district court, the transfer rule is simply not subject to antitrust analysis because it is beyond the scope of the Act.

Of course, “[o]ur review is not limited to a consideration of the grounds upon which the district court decided the issues; we can affirm the district court on any grounds supported by the record.” Weiser v. United States, 959 F.2d 146, 147 (9th Cir.1992). We need not reach the difficult issue of whether collegiate athletic association eligibility rules such as the Pac-10 transfer rule do not involve commercial activity and hence are immune from Sherman Act scrutiny. For purposes of our analysis, we assume, without deciding, that the transfer rule is subject to the federal antitrust laws.

B

Tanaka does not contend that the transfer rule is unlawful per se, but rather concedes that it is subject to rule of reason analysis. We agree. See NCAA v. Bd. of Regents, 468 U.S. 85, 117 , 104 S.Ct. 2948 , 82 L.Ed.2d 70 (1984) (holding that NCAA restrictions on the televising of college football games, which involved limits on output and price-fixing, was subject to rule of reason analysis); Mackey v. Nat’l Football League, 543 F.2d 606 (8th Cir.1976) (holding that league’s restrictions on free agency were subject to rule of reason analysis despite restrictions’ resemblance to restraints ordinarily constituting per se antitrust violations, such as group boycotts and concerted refusals to deal).

A restraint violates the rule of reason if the restraint’s harm to competition outweighs its procompetitive effects. Hairston, 101 F.3d at 1319 . The plaintiff bears the initial burden of showing that the restraint produces “significant anti-competitive effects” within a “relevant market.” Id. If the plaintiff meets this burden, the defendant must come forward with evidence of the restraint’s procompetitive effects. The plaintiff must then show that “any legitimate objectives can be achieved in a substantially less restrictive manner.” Id.

We have explained that the term “relevant market”

Oltz v. St. Peter’s Cmty. Hosp., 861 F.2d 1440, 1446 (9th Cir.1988) (quoting Moore v. Jas. H. Matthews & Co., 550 F.2d 1207, 1218 (9th Cir.1977)) (alteration in original) (internal citations omitted). Failure to identify a relevant market is a proper ground for dismissing a Sherman Act claim. Big Bear Lodging Ass’n v. Snow Summit, Inc., 182 F.3d 1096, 1105 (9th Cir.1999).

1

Tanaka’s complaint alleges that the relevant geographic market is Los Angeles and the relevant product market is the “UCLA women’s soccer program.” Neither of these “markets” is appropriately defined for antitrust purposes, even at this stage of the litigation. First, Tanaka fails to allege that Los Angeles is an “area of effective competition” for student-athletes competing for positions in women’s intercollegiate soccer programs. In her reply brief, Tanaka explains that “the relevant market is [ ] Los Angeles because [she] wanted to be close to her family.” Of course, Tanaka’s personal preference to remain in the Los Angeles area is irrelevant to the question of whether Los Angeles is an area of effective competition for the services of women’s intercollegiate soccer players. Cf. id. (plaintiffs’ complaint identifying Big Bear Valley as the relevant market fails as a matter of law where plaintiffs did not allege that “Big Bear Valley is the area of effective competition in which buyers of these products can find alternative sources of supply”). In fact, according to her own allegations, Tanaka herself was “heavily recruited by universities across the country,” from Connecticut to Washington. Tanaka’s own experience strongly suggests that the relevant geographic market is national in scope.

Second, Tanaka has failed to identify an appropriately defined product market. Her conclusory assertion that the “UCLA women’s soccer program” is “unique” and hence “not interchangeable with any other program in Los Angeles” is insufficient. Again, Tanaka’s strictly personal preference to remain in Los Angeles is irrelevant to the antitrust inquiry before us. Moreover, her limitation of the relevant product market to a single athletic program is especially unavailing insofar as the very existence of any given intercollegiate ath letic program is predicated upon the existence of a field of competition composed of other, similar programs. Such programs compete in the recruiting of student-athletes and, hence, are interchangeable with each other for antitrust purposes. Since Tanaka herself was recruited by numerous Pac-10 and non-Pac-10 programs located throughout the country, nothing beyond her personal preferences suggests that UCLA was the only potential option for women’s intercollegiate soccer players in her position.

Finally, Tanaka has failed to allege that the transfer rule has had significant anti-competitive effects within a relevant market, however defined. If the relevant market is national in scope, as Tanaka’s own complaint suggests, the transfer rule most certainly does not have a significant anti-competitive effect. By its own terms, the Pac-10 transfer rule applies only to intra-conference transfers; it has no application to student-athletes who transfer to nonmember institutions. If, on the other hand, Tanaka were to amend her complaint and attempt to identify the Pac-10 as the relevant market, she would still be unable to allege that the transfer rule has had a significant anticompetitive effect on this “market” given her assertion that she “is the only transferring athlete who had the sanctions imposed on her” (emphasis in original), and further that “[i]n all other instances where student-athletes transferred from USC, Pacific-10 penalties had not been used.”

In fact, Tanaka depicts USC’s invocation of transfer rule sanctions against her as an isolated act of retaliation. While “hundreds of student-athletes transfer each year” without sanction, Tanaka alleges that she was “singled out” in an attempt by USC to “keep her at the school to ‘oversee’ any testimony that she might give” concerning the alleged incident of academic fraud. In short, Tanaka alleges nothing more than a personal injury to herself, not an injury to a definable market. But the antitrust laws “were enacted for ‘the protection of competition, not competitors.’ ” Atlantic Richfield Co. v. USA Petroleum Co., 495 U.S. 328, 338 , 110 S.Ct. 1884 , 109 L.Ed.2d 333 (1990) (quoting Brown Shoe Co. v. United States, 370 U.S. 294, 320 , 82 S.Ct. 1502 , 8 L.Ed.2d 510 (1962)). As we have explained, “[i]t is the impact upon competitive conditions in a definable market which distinguishes the antitrust violation from the ordinary business tort. [The] failure to allege injury to competition is a proper ground for dismissal by judgment on the pleadings.” McGlinchy v. Shell Chem. Co., 845 F.2d 802 , 812-13 (9th Cir.1988); see also Adaptive Power Solutions, LLC v. Hughes Missile Sys. Co., 141 F.3d 947 , 952 (9th Cir.1998) (holding that agreement between missile component market’s only two buyers not to purchase missile component from one of the market’s only two sellers survived rule of reason analysis; although defendants injured seller by driving it out of market, there was no injury to competition, only a temporary decline in number of competitors).

2

We note that, in theory, restrictions on student-athlete transfers could be loosely analogized to the National Football League free agency restrictions which the Eighth Circuit concluded failed rule of reason analysis in Mackey . At issue in Mack-ey was the so-called “Rozelle Rule,” which granted the league commissioner the authority to award to a team that lost a free agent one or more players from the acquiring team as compensation, thus restraining the market for players’ services. 543 F.2d at 620 . Nevertheless, unlike the Pac-10 transfer rule, the “Rozelle Rule applie[d] to every NFL player regardless of his status or ability ... [and was] unlim ited in duration ... operating] as a perpetual restriction on a player’s ability to sell his services in an open market throughout his career.” Id. at 622 . Even if some student-athlete transfer restrictions could be characterized as “concerted refusals to deal” having the effect of locking players in their current athletic programs, and hence roughly analogous to the Rozelle Rule, this would be of no avail to Tanaka, given her failure to allege that the Pac-10 transfer rule has had a significant anticompetitive effect within a properly defined geographic and product market, one that is not circumscribed by Tanaka’s own (irrelevant) personal preferences.

Tanaka’s reliance upon our decision in Hairston is also misplaced. In Hairston , we considered a Sherman Act challenge to a Pac-10 agreement to sanction the University of Washington’s football program for recruiting violations. While affirming the district court’s grant of summary judgment in favor of the Pac-10 under a rule of reason analysis, we remarked, without further explanation, that the plaintiffs “met their initial burden [of demonstrating significant anticompetitive effects] by showing that the Pac-10 members banned UW from participating in bowl games for two years.” Hairston, 101 F.3d at 1319 . We hesitate to read too much into this unexplained remark; certainly, it does not remotely suggest that the mere exclusion of one competitor from a market, without more, satisfies the plaintiffs burden to show harm to a relevant market. It is simply unclear what additional facts the court in Hairston relied on to conclude that the plaintiffs met their initial burden. In any event, the plaintiffs in Hairston did not, as Tanaka has done, allege that the University of Washington was the only competitor against whom the Pac-10 recruiting rule at issue was ever applied.

C

By attempting to restrict the relevant market to a single athletic program in Los Angeles based solely on her own preferences, Tanaka has failed to identify a relevant market for antitrust purposes. If the relevant market is national in scope, the Pac-10 transfer rule could not have a significant anticompetitive effect, because by its own terms the rule does not apply to interconference transfers. Moreover, even if the relevant market is limited to the Pac-10 itself, Tanaka characterizes the Pac-10’s imposition of sanctions against her for her intraconference transfer as an isolated act of retaliation. Tanaka simply has no antitrust cause of action; thus, it is “clear that [her] complaint cannot be saved by further amendment.” Big Bear Lodging Ass’n, 182 F.3d at 1101 .

Ill

For the foregoing reasons, the judgment of the district court is

AFFIRMED.

. The Pac-10's other members are Arizona, Arizona State, Oregon, Oregon State, Stanford, University of California, Berkeley, UCLA, Washington and Washington State. . A student-athlete is eligible to participate in Pac-10 competition for a maximum of four years and has five years to complete this eligibility. . Tanaka’s Sherman Act claim against the NCAA fails for the additional reason that she has not alleged that the NCAA actually “contracted, combined or conspired” with the other defendants. Indeed, the NCAA’s own transfer rule was not applied to Tanaka. . The district court did not abuse its discretion in dismissing Tanaka's state breach of contract claims without prejudice pursuant to 28 U.S.C. § 1367 (c)(3), which provides that “district courts may decline to exercise supplemental jurisdiction over a claim ... if ... (3) the district court has dismissed all claims over which it has original jurisdiction.” 28 U.S.C. § 1367 (c)(3).