Section 505 of the Copyright Act provides that a district court "may ... award a reasonable attorney's fee to the prevailing party." 17 U.S.C. § 505 . The question presented here is whether a court, in exercising that authority, should give substantial weight to the objective reasonableness of the losing party's position. The answer, as both decisions below held, is yes-the court should. But the court must also give due consideration to all other circumstances relevant to granting fees; and it retains discretion, in light of those factors, to make an award even when the losing party advanced a reasonable claim or defense. Because we are not certain that the lower courts here understood the full scope of that discretion, we return the case for further consideration of the prevailing party's fee application.

I

Petitioner Supap Kirtsaeng, a citizen of Thailand, came to the United States 20 years ago to study math at Cornell University. He quickly figured out that respondent John Wiley & Sons, an academic publishing company, sold virtually identical English-language textbooks in the two countries-but for far less in Thailand than in the United States. Seeing a ripe opportunity for arbitrage, Kirtsaeng asked family and friends to buy the foreign editions in Thai bookstores and ship them to him in New York. He then resold the textbooks to American students, reimbursed his Thai suppliers, and pocketed a tidy profit.

Wiley sued Kirtsaeng for copyright infringement, claiming that his activities violated *1984 its exclusive right to distribute the textbooks. See 17 U.S.C. §§ 106 (3), 602(a)(1). Kirtsaeng invoked the "first-sale doctrine" as a defense. That doctrine typically enables the lawful owner of a book (or other work) to resell or otherwise dispose of it as he wishes. See § 109(a). But Wiley contended that the first-sale doctrine did not apply when a book (like those Kirtsaeng sold) was manufactured abroad.

At the time, courts were in conflict on that issue. Some thought, as Kirtsaeng did, that the first-sale doctrine permitted the resale of foreign-made books; others maintained, along with Wiley, that it did not. And this Court, in its first pass at the issue, divided 4 to 4. See Costco Wholesale Corp. v. Omega, S.A., 562 U.S. 40 , 131 S.Ct. 565 , 178 L.Ed.2d 470 (2010) ( per curiam ). In this case, the District Court sided with Wiley; so too did a divided panel of the Court of Appeals for the Second Circuit. See 654 F.3d 210 , 214, 222 (2011). To settle the continuing conflict, this Court granted Kirtsaeng's petition for certiorari and reversed the Second Circuit in a 6-to-3 decision, thus establishing that the first-sale doctrine allows the resale of foreign-made books, just as it does domestic ones. See Kirtsaeng v. John Wiley & Sons, Inc., 568 U.S. ----, ----, 133 S.Ct. 1351 , 1355-1356, 185 L.Ed.2d 392 (2013).

Returning victorious to the District Court, Kirtsaeng invoked § 505 to seek more than $2 million in attorney's fees from Wiley. The court denied his motion. Relying on Second Circuit precedent, the court gave "substantial weight" to the "objective reasonableness" of Wiley's infringement claim. See No. 08-cv-07834 (S.D.N.Y., Dec. 20, 2013), App. to Pet. for Cert. 18a, 2013 WL 6722887 , *4. In explanation of that approach, the court stated that "the imposition of a fee award against a copyright holder with an objectively reasonable"-although unsuccessful-"litigation position will generally not promote the purposes of the Copyright Act." Id., at 11a (quoting Matthew Bender & Co. v. West Publishing Co., 240 F.3d 116 , 122 (C.A.2 2001) (emphasis deleted)). Here, Wiley's position was reasonable: After all, several Courts of Appeals and three Justices of the Supreme Court had agreed with it. See App. to Pet. for Cert. 12a. And according to the District Court, no other circumstance "overr[o]de" that objective reasonableness, so as to warrant fee-shifting. Id., at 22a. The Court of Appeals affirmed, concluding in a brief summary order that "the district court properly placed 'substantial weight' on the reasonableness of [Wiley's] position" and committed no abuse of discretion in deciding that other "factors did not outweigh" the reasonableness finding. 605 Fed.Appx. 48 , 49, 50 (C.A.2 2015).

We granted certiorari, 577 U.S. ----, 136 S.Ct. 890 , 193 L.Ed.2d 783 (2016), to resolve disagreement in the lower courts about how to address an application for attorney's fees in a copyright case. 1

II

Section 505 states that a district court "may ... award a reasonable attorney's fee to the prevailing party." It thus authorizes fee-shifting, but without specifying standards that courts should adopt, or *1985 guideposts they should use, in determining when such awards are appropriate.

In Fogerty v. Fantasy, Inc., 510 U.S. 517 , 114 S.Ct. 1023 , 127 L.Ed.2d 455 (1994), this Court recognized the broad leeway § 505 gives to district courts-but also established several principles and criteria to guide their decisions. See id., at 519 , 114 S.Ct. 1023 (asking "what standards should inform" the exercise of the trial court's authority). The statutory language, we stated, "clearly connotes discretion," and eschews any "precise rule or formula" for awarding fees. Id., at 533, 534 , 114 S.Ct. 1023 . Still, we established a pair of restrictions. First, a district court may not "award[ ] attorney's fees as a matter of course"; rather, a court must make a more particularized, case-by-case assessment. Id., at 533 , 114 S.Ct. 1023 . Second, a court may not treat prevailing plaintiffs and prevailing defendants any differently; defendants should be "encouraged to litigate [meritorious copyright defenses] to the same extent that plaintiffs are encouraged to litigate meritorious claims of infringement." Id., at 527 , 114 S.Ct. 1023 . In addition, we noted with approval "several nonexclusive factors" to inform a court's fee-shifting decisions: "frivolousness, motivation, objective unreasonableness[,] and the need in particular circumstances to advance considerations of compensation and deterrence." Id., at 534, n. 19 , 114 S.Ct. 1023 . And we left open the possibility of providing further guidance in the future, in response to (and grounded on) lower courts' evolving experience. See id., at 534-535 , 114 S.Ct. 1023 ; Martin v. Franklin Capital Corp., 546 U.S. 132 , 140, n., 126 S.Ct. 704 , 163 L.Ed.2d 547 (2005) (noting that Fogerty was not intended to be the end of the matter).

The parties here, though sharing some common ground, now dispute what else we should say to district courts. Both Kirtsaeng and Wiley agree-as they must-that § 505 grants courts wide latitude to award attorney's fees based on the totality of circumstances in a case. See Brief for Petitioner 17; Brief for Respondent 35. Yet both reject the position, taken by some Courts of Appeals, see supra, at 1984, n. 1, that Fogerty spelled out the only appropriate limits on judicial discretion-in other words, that each district court should otherwise proceed as it sees fit, assigning whatever weight to whatever factors it chooses. Rather, Kirtsaeng and Wiley both call, in almost identical language, for "[c]hanneling district court discretion towards the purposes of the Copyright Act." Brief for Petitioner 16; see Brief for Respondent 21 ("[A]n appellate court [should] channel a district court's discretion so that it ... further[s] the goals of the Copyright Act"). (And indeed, as discussed later, both describe those purposes identically. See infra, at 1986.) But at that point, the two part ways. Wiley argues that giving substantial weight to the reasonableness of a losing party's position will best serve the Act's objectives. See Brief for Respondent 24-35. By contrast, Kirtsaeng favors giving special consideration to whether a lawsuit resolved an important and close legal issue and thus "meaningfully clarifie[d]" copyright law. Brief for Petitioner 36; see id., at 41-44.

We join both parties in seeing a need for some additional guidance respecting the application of § 505. In addressing other open-ended fee-shifting statutes, this Court has emphasized that "in a system of laws discretion is rarely without limits." Flight Attendants v. Zipes, 491 U.S. 754 , 758, 109 S.Ct. 2732 , 105 L.Ed.2d 639 (1989) ; see *1986 Halo Electronics, Inc. v. Pulse Electronics, Inc., --- U.S. ----, ----, 136 S.Ct. 1923 , 1931, --- L.Ed.2d ----, 2016 WL 3221515 (2016) ante, at 8. Without governing standards or principles, such provisions threaten to condone judicial "whim" or predilection. Martin, 546 U.S., at 139 , 126 S.Ct. 704 ; see also ibid. ("[A] motion to [a court's] discretion is a motion, not to its inclination, but to its judgment; and its judgment is to be guided by sound legal principles" (quoting United States v. Burr, 25 F.Cas. 30 , 35 (No. 14,692d) (C.C.Va. 1807) (Marshall, C. J.))). At the least, utterly freewheeling inquiries often deprive litigants of "the basic principle of justice that like cases should be decided alike," Martin, 546 U.S., at 139 , 126 S.Ct. 704 -as when, for example, one judge thinks the parties' "motivation [s]" determinative and another believes the need for "compensation" trumps all else, Fogerty, 510 U.S., at 534, n. 19 , 114 S.Ct. 1023 . And so too, such unconstrained discretion prevents individuals from predicting how fee decisions will turn out, and thus from making properly informed judgments about whether to litigate. For those reasons, when applying fee-shifting laws with "no explicit limit or condition," Halo, --- U.S. at, ----, --- S.Ct., at --- -, ante, at 8, we have nonetheless "found limits" in them-and we have done so, just as both parties urge, by looking to "the large objectives of the relevant Act," Zipes, 491 U.S., at 759 , 109 S.Ct. 2732 (internal quotation marks omitted); see supra, at 1985 - 1986.

In accord with such precedents, we must consider if either Wiley's or Kirtsaeng's proposal well advances the Copyright Act's goals. Those objectives are well settled. As Fogerty explained, "copyright law ultimately serves the purpose of enriching the general public through access to creative works." 510 U.S., at 527 , 114 S.Ct. 1023 ; see U.S. Const., Art. I, § 8, cl. 8 ("To promote the Progress of Science and useful Arts"). The statute achieves that end by striking a balance between two subsidiary aims: encouraging and rewarding authors' creations while also enabling others to build on that work. See Fogerty, 510 U.S., at 526 , 114 S.Ct. 1023 . Accordingly, fee awards under § 505 should encourage the types of lawsuits that promote those purposes. (That is why, for example, Fogerty insisted on treating prevailing plaintiffs and prevailing defendants alike-because the one could "further the policies of the Copyright Act every bit as much as" the other. 510 U.S., at 527 , 114 S.Ct. 1023 . ) On that much, both parties agree. Brief for Petitioner 37; Brief for Respondent 29-30. The contested issue is whether giving substantial weight to the objective (un)reasonableness of a losing party's litigating position-or, alternatively, to a lawsuit's role in settling significant and uncertain legal issues-will predictably encourage such useful copyright litigation.

The objective-reasonableness approach that Wiley favors passes that test because it both encourages parties with strong legal positions to stand on their rights and deters those with weak ones from proceeding with litigation. When a litigant-whether plaintiff or defendant-is clearly correct, the likelihood that he will recover fees from the opposing ( i.e., unreasonable) party gives him an incentive to litigate the case all the way to the end. The holder of a copyright that has obviously been infringed has good reason to bring and maintain a suit even if the damages at stake are small; and likewise, a person defending against a patently meritless copyright claim has every incentive to keep fighting, no matter that attorney's fees in a protracted suit might be as or more costly than a settlement. Conversely, when a person (again, whether plaintiff or defendant) has an unreasonable litigating position, the likelihood that he will have to pay two sets of fees discourages *1987 legal action. The copyright holder with no reasonable infringement claim has good reason not to bring suit in the first instance (knowing he cannot force a settlement and will have to proceed to judgment); and the infringer with no reasonable defense has every reason to give in quickly, before each side's litigation costs mount. All of those results promote the Copyright Act's purposes, by enhancing the probability that both creators and users ( i.e., potential plaintiffs and defendants) will enjoy the substantive rights the statute provides.

By contrast, Kirtsaeng's proposal would not produce any sure benefits. We accept his premise that litigation of close cases can help ensure that "the boundaries of copyright law [are] demarcated as clearly as possible," thus advancing the public interest in creative work. Brief for Petitioner 19 (quoting Fogerty, 510 U.S., at 527 , 114 S.Ct. 1023 ). But we cannot agree that fee-shifting will necessarily, or even usually, encourage parties to litigate those cases to judgment. Fee awards are a double-edged sword: They increase the reward for a victory-but also enhance the penalty for a defeat. And the hallmark of hard cases is that no party can be confident if he will win or lose. That means Kirtsaeng's approach could just as easily discourage as encourage parties to pursue the kinds of suits that "meaningfully clarif[y]" copyright law. Brief for Petitioner 36. It would (by definition) raise the stakes of such suits; but whether those higher stakes would provide an incentive-or instead a disincentive-to litigate hinges on a party's attitude toward risk. Is the person risk-preferring or risk-averse-a high-roller or a penny-ante type? Only the former would litigate more in Kirtsaeng's world. See Posner, An Economic Approach to Legal Procedure and Judicial Administration, 2 J. Legal Studies 399, 428 (1973) (fees "make[ ] the expected value of litigation less for risk-averse litigants, which will encourage [them to] settle[ ]"). And Kirtsaeng offers no reason to think that serious gamblers predominate. See, e.g., Texas Industries, Inc. v. Radcliff Materials, Inc., 451 U.S. 630 , 636, n. 8, 101 S.Ct. 2061 , 68 L.Ed.2d 500 (1981) ("Economists disagree over whether business decisionmakers[ ] are 'risk averse' "); CIGNA Corp. v. Amara, 563 U.S. 421 , 430, 131 S.Ct. 1866 , 179 L.Ed.2d 843 (2011) ("[M]ost individuals are risk averse"). So the value of his standard, unlike Wiley's, is entirely speculative. 2

What is more, Wiley's approach is more administrable than Kirtsaeng's. A district court that has ruled on the merits of a copyright case can easily assess whether the losing party advanced an unreasonable claim or defense. That is closely related to what the court has already done: In deciding any case, a judge cannot help but consider the strength and weakness of each side's arguments. By contrast, a judge may not know at the conclusion of a *1988 suit whether a newly decided issue will have, as Kirtsaeng thinks critical, broad legal significance. The precedent-setting, law-clarifying value of a decision may become apparent only in retrospect-sometimes, not until many years later. And so too a decision's practical impact (to the extent Kirtsaeng would have courts separately consider that factor). District courts are not accustomed to evaluating in real time either the jurisprudential or the on-the-ground import of their rulings. Exactly how they would do so is uncertain (Kirtsaeng points to no other context in which courts undertake such an analysis), but we fear that the inquiry would implicate our oft-stated concern that an application for attorney's fees "should not result in a second major litigation." Zipes, 491 U.S., at 766 , 109 S.Ct. 2732 (quoting Hensley v. Eckerhart, 461 U.S. 424 , 437, 103 S.Ct. 1933 , 76 L.Ed.2d 40 (1983) ). And we suspect that even at the end of that post-lawsuit lawsuit, the results would typically reflect little more than educated guesses.

Contrary to Kirtsaeng's view, placing substantial weight on objective reasonableness also treats plaintiffs and defendants even-handedly, as Fogerty commands. No matter which side wins a case, the court must assess whether the other side's position was (un)reasonable. And of course, both plaintiffs and defendants can (and sometimes do) make unreasonable arguments. Kirtsaeng claims that the reasonableness inquiry systematically favors plaintiffs because a losing defendant "will virtually always be found to have done something culpable." Brief for Petitioner 29 (emphasis in original). But that conflates two different questions: whether a defendant in fact infringed a copyright and whether he made serious arguments in defense of his conduct. Courts every day see reasonable defenses that ultimately fail (just as they see reasonable claims that come to nothing); in this context, as in any other, they are capable of distinguishing between those defenses (or claims) and the objectively unreasonable variety. And if some court confuses the issue of liability with that of reasonableness, its fee award should be reversed for abuse of discretion. 3

All of that said, objective reasonableness can be only an important factor in assessing fee applications-not the controlling one. As we recognized in Fogerty, § 505 confers broad discretion on district courts and, in deciding whether to fee-shift, they must take into account a range of considerations beyond the reasonableness of litigating positions. See supra, at 1985. That means in any given case a court may award fees even though the losing party offered reasonable arguments (or, conversely, deny fees even though the losing party made unreasonable ones). For example, a court may order fee-shifting *1989 because of a party's litigation misconduct, whatever the reasonableness of his claims or defenses. See, e.g., Viva Video, Inc. v. Cabrera, 9 Fed.Appx. 77 , 80 (C.A.2 2001). Or a court may do so to deter repeated instances of copyright infringement or overaggressive assertions of copyright claims, again even if the losing position was reasonable in a particular case. See, e.g., Bridgeport Music, Inc. v. WB Music Corp., 520 F.3d 588 , 593-595 (C.A.6 2008) (awarding fees against a copyright holder who filed hundreds of suits on an overbroad legal theory, including in a subset of cases in which it was objectively reasonable). Although objective reasonableness carries significant weight, courts must view all the circumstances of a case on their own terms, in light of the Copyright Act's essential goals.

And on that score, Kirtsaeng has raised serious questions about how fee-shifting actually operates in the Second Circuit. To be sure, the Court of Appeals' framing of the inquiry resembles our own: It calls for a district court to give "substantial weight" to the reasonableness of a losing party's litigating positions while also considering other relevant circumstances. See 605 Fed.Appx., at 49-50 ; Matthew Bender, 240 F.3d, at 122 . But the Court of Appeals' language at times suggests that a finding of reasonableness raises a presumption against granting fees, see ibid. ; supra, at 1983 - 1985-and that goes too far in cabining how a district court must structure its analysis and what it may conclude from its review of relevant factors. Still more, district courts in the Second Circuit appear to have overly learned the Court of Appeals' lesson, turning "substantial" into more nearly "dispositive" weight. As Kirtsaeng notes, hardly any decisions in that Circuit have granted fees when the losing party raised a reasonable argument (and none have denied fees when the losing party failed to do so). See Reply Brief 15. For these reasons, we vacate the decision below so that the District Court can take another look at Kirtsaeng's fee application. In sending back the case for this purpose, we do not at all intimate that the District Court should reach a different conclusion. Rather, we merely ensure that the court will evaluate the motion consistent with the analysis we have set out-giving substantial weight to the reasonableness of Wiley's litigating position, but also taking into account all other relevant factors.

* * *

The judgment of the Court of Appeals is vacated, and the case is remanded for further proceedings consistent with this opinion.

It is so ordered.