OPINION OF THE COURT

In this appeal, we consider whether a federal district court properly relied on Federal Rule of Civil Procedure 60(b)(6) to vacate a default judgment entered by another district court. We conclude that it did not, and we will remand so that it may consider whether to set aside the default judgment under Federal Rule of Civil Procedure 60(b)(4).

I.

Budget Blinds, Inc. (“BBI”) is a California corporation that franchises mobile window covering businesses throughout the United States. According to an affidavit that BBI’s Chief Operating Officer filed with the District Court for the Central District of California, BBI was founded in 1992 and had about 800 territories and 570 licensees nationwide as of October 17, 2005, the date of the affidavit. BBI owns and licenses two trademarks that it has registered with the Principal Register of the United States Patent and Trademark Office: (1) the name “Budget Blinds” (registered on December 21, 1993); and (2) a service mark consisting of the words “Budget Blinds” in a specific font and configuration (registered on February 18, 2003).

Valerie White owns a New Jersey corporation called “Val U Blinds, Inc.” that focuses on the design and installation of window blinds. According to an affidavit that White filed with the District Court for the District of New Jersey, she operates this business from a home office in her basement and garage, and her sales are limited to areas of New Jersey. The affidavit further states that she conducted business from 1988 to June 1, 2004 as “Budget Blinds” or “Budget Blinds of NJ,” and that she registered her business as a New Jersey domestic corporation in Sep*247tember 1997 under the name “Budget Blinds of NJ, Inc.” She changed her business’s name to “Val U Blinds” pursuant to the Settlement Agreement that we describe below.

In a letter to Valerie White dated November 25, 2003, BBI’s Legal Manager stated that White’s use of the name “Budget Blinds” was a violation of BBI’s “federal, state, and common law trademark rights” and was likely to cause public confusion about the origin of White’s goods. White’s attorney Ronald J. Nelson responded in a letter dated December 1, 2003 that White had used the name “Budget Blinds” since 1988, prior to BBI’s first use of the name. Nelson’s letter added that White had established “Budget Blinds” as a common law trademark in the New Jersey counties served by her business and that BBI’s franchisees in this area were infringing her rights by using the name. In a letter to Nelson dated February 12, 2004, BBI’s counsel questioned the existence of a common law trademark and stated that “BBI is prepared to bring a lawsuit in California pursuant to the Lanham Act to enjoin your client’s infringing activities” if the parties could not reach a mutually-agreeable resolution.

After several additional communications, the parties entered a Settlement Agreement on April 14, 2004. Under the Agreement, BBI would pay White and Budget Blinds of NJ, Inc. (referred to as “the Corporation” in the Agreement) $160,000 “for the purchase and transfer of any and all interest, claim, or ownership in or relating to the trade name and service mark ‘Budget Blinds’ ” and any confusingly similar names or marks. Section 4 of the Agreement provided that after June 1, 2004, White and the Corporation “shall not operate or do business” under the trade name and service mark “Budget Blinds” “or any other name or in any manner that might tend to give the general public the impression that White [or her business] is in any way associated or affiliated with BBI, or any of the businesses conducted by it or other franchisees or licensees of the Marks.” Among other things, Section 4 required White to change her existing corporation’s name from “Budget Blinds of New Jersey, Inc.” to “BB of NJ, Inc.,” to conduct all future business using a new corporation to be established under the name “Val U Blinds, Inc.,” and to remove the “Budget Blinds” trade name and mark, as well as confusingly similar names or marks, from her company’s advertising, signs, letterheads, stationery, printed matter, and other forms. Most importantly for purposes of this litigation, Section 4(f) of the Agreement also instructed White to take specific steps to disassociate her company’s telephone number from the name “Budget Blinds.” At the time of the Agreement, the Yellow Pages directories for Burlington, Gloucester, and Camden Counties listed a phone number for White’s company next to the name “Budget Blinds.”1 To address this situation, Section 4(f) of the Agreement provided in relevant part:

Promptly after execution of this Agreement, White, the Corporation and Val-U Blinds shall direct Verizon ... in writing, with a copy to counsel for BBI ..., (i) that all of the said three advertisements shall not be renewed in the said three county editions, or elsewhere; and (ii) that customers calling directory as*248sistance in any of those three counties (or anywhere else) on or after June 1, 2004, should no longer be given [the phone number for White’s company] or any other number related to or affiliated with White, the Corporation, or Yal-U Blinds in response to an inquiry for the telephone number of “Budget Blinds.” From time-to-time thereafter, upon the reasonable request of BBI, White, the Corporation and Val-U Blinds shall provide similar written directions to other directory publishers (including publishers of Internet-based “directories”) identified by BBI.

In addition to these substantive provisions, the Agreement stated that White and her corporations “hereby irrevocably appoint BBI as their respective lawful attorney-in-fact with authority to file any document in the name of and on behalf of White, the Corporation or Yal-U Blinds for the purpose of taking any of the actions required by this Section 4 upon the event of a Default.” Finally, the Agreement contained a choice-of-law clause: “This Settlement Agreement will be governed by and construed under the laws of the State of California.”

White says that she made a good-faith effort to comply with her obligations under the Agreement, including those in Section 4(f) related to her telephone number. The record contains a letter dated May 5, 2004 from White’s attorney to Verizon Customer Service directing Verizon “not to renew or republish their existing advertisements in all Yellow Pages (including Burlington, Gloucester, Camden and Middlesex County editions) in which they have used the phrase ‘Budget Blinds’ in whole or in part.” The letter also directs Verizon “to instruct Directory Assistance, beginning June 1, 2004, to answer all inquiries for ‘Budget Blinds’ by not providing the telephone numbers of my client....” The record includes similar letters addressed to YellowPages.com and Verizon.com, both of which are dated June 4, 2004.

Nonetheless, Verizon reprinted the listing that provided the phone number for White’s company under the name “Budget Blinds” in the updated print editions of its Yellow Pages directories for Burlington, Gloucester, and Camden Counties and the corresponding online directories. In a letter dated February 1, 2005, BBI’s counsel informed White’s counsel that the continuing existence of these listings was a violation of the Agreement, and that “it appears that the only way to remedy this violation is for your client to assign to BBI or to a franchisee designated by BBI” the telephone number for White’s company. In response, White’s counsel proposed setting up a recording that would give callers to this telephone number a choice between connecting to White’s company or to the local BBI franchisee. BBI rejected this proposal in a letter dated February 23, 2005, reiterating its demand that White assign the number to BBI and stating that it would deem White and her company to be in breach of the Agreement “as long as there is any circumstance under which a person looking in a current telephone book and calling a number listed under the name ‘Budget Blinds’ will reach your clients.” In a letter dated March 9, 2005, White’s counsel informed BBI that, although White would do everything that she considered “reasonably possible to mitigate fully the effects of the unauthorized re-printing,” she would not surrender the phone number itself.

BBI filed a Complaint against White, Val U Blinds, Inc., and Budget Blinds of NJ, Inc.2 (“Defendants”), which was dock*249eted on April 7, 2005 in the United States District Court for the Central District of California, Southern Division. The Complaint lists eight causes of action: (1) breach of contract; (2) specific performance; (3) federal trademark infringement; (4) state trademark infringement; (5) federal false designation of origin; (6) federal trademark dilution; (7) state trademark dilution; and (8) violation of the California Unfair Competition Act. As relief, BBI requested an injunction against further violation of the agreement,3 actual and punitive damages, an accounting of defendants’ profits, treble damages, costs of suit, attorneys’ fees, and pre-judgment interest. The Complaint asserts that venue is proper in the Central District of California “because a substantial part of the events giving rise to Budget Blinds’ claims occurred within this District in that the harm to Budget Blinds has occurred in this district and the Defendants have intentionally directed their actions toward Budget Blinds, which is headquartered in this District.” The Defendants did not file a response to the Complaint even though BBI sent them notice of it. White’s counsel characterizes this as an intentional decision, motivated in part by White’s limited resources, to ignore the Complaint on the ground that the California court allegedly lacked personal jurisdiction over the Defendants.

On October 7, 2005, the District Court for the Central District of California entered a default judgment against the Defendants, granting BBI the injunctive relief that it requested in its Complaint and $83,083.51 in monetary relief.4 BBI registered this default judgment in the District Court for the District of New Jersey on November 22, 2005. On January 20, 2006, BBI gave the Defendants notice that it would file a motion on February 14 in the District Court for the District of New Jersey for an order directing the turnover of funds from White’s checking account to BBI in the amount specified in the judgment. On January 24, the Defendants gave notice that they would file a cross-motion in the New Jersey District Court to vacate the default judgment, and they filed a Brief in opposition to the turnover motion and in support of the cross-motion. In this Brief, the Defendants asserted that the default judgment should be declared null and void because the California district court had lacked personal jurisdiction over them. In support of this Brief, White submitted an affidavit explaining inter alia that neither she nor her company had ever physically entered California, owned property in California, solicited business in California, purchased or sold any goods in California, or even maintained an internet website. Both the Brief and White’s affidavit asserted that an earlier draft of the Agreement contained a provision granting White’s consent to personal jurisdiction in the Central District of California but that the Defendants refused to allow inclusion of this provision in the final Agreement. BBI filed a Memorandum of Law on Janu*250ary 31 opposing the Defendants’ cross-motion to vacate the default judgment and arguing that the California court had personal jurisdiction over the Defendants.

On April 5, 2006, the District Court for the District of New Jersey issued an Opinion and Order in response to the parties’ motions. See Budget Blinds v. White, No. 05-CV-388, 2006 U.S. Dist. LEXIS 17207, 2006 WL 891187 (D.N.J. Apr.5, 2006). Instead of resolving the personal jurisdiction issue, the District Court vacated the default judgment under Federal Rule of Civil Procedure 60(b)(6), which provides: “On motion and just terms, the court may relieve a party or its legal representative from a final judgment, order, or proceeding for ... any other reason that justifies relief.” To decide the Rule 60(b)(6) issue, the District Court employed a three-part test derived from Harad v. Aetna Casualty & Surety Co., 839 F.2d 979 (3d Cir. 1988): “When determining whether to vacate a default judgment, a court should consider the following factors: (1) whether the defendant has a meritorious defense; (2) whether culpable conduct of the defendant led to the default; and (3) whether the plaintiff will be prejudiced.” Budget Blinds, 2006 U.S. Dist. LEXIS 17207, at *4, 2006 WL 891187, at *1 (citing Harad, 839 F.2d at 982). First, the District Court found that the Defendants had an “extremely” meritorious defense, stating: “The facts establish that White fully complied with her obligations under the settlement agreement and therefore the provision [in the Agreement] staying all litigation is, most likely, still in effect.” 2006 U.S. Dist. LEXIS 17207, at *5, 2006 WL 891187, at *2. Second, the District Court concluded that the culpability factor “arguably benefits either party” because White had “more than a colorable argument that the district court lacked personal jurisdiction” but it would have been “prudent” for her to appear before the California court to make this argument. Id. at *5-*6 & n. 2, 2006 WL 891187, at 2 & n. 2. Third, the District Court found that “BBI will suffer absolutely no prejudice should the Court order vacatur.” Id. at *7, 2006 WL 891187, at *2. Weighing these three factors, the District Court concluded that va-catur was appropriate. Id. at *7-*8, 2006 WL 891187, at *2.

Even though the briefs of both parties had focused on personal jurisdiction, the New Jersey District Court did not decide this issue, explaining in a footnote that “[t]he Court need not pass on whether the California district court actually possessed personal jurisdiction over White for purposes of a Rule 60(b)(6) inquiry. It is enough to know that White possessed more than a colorable argument that the district court lacked personal jurisdiction.” Id. at *6 n. 2, 2006 WL 891187, at *2 n. 2. The District Court did, however, “grant BBI’s reasonable legal fees and costs associated with obtaining the default judgment and registering the judgment in this Court,” in part because “a prudent attorney” would have responded to BBI’s complaint despite his or her belief that the California district court lacked personal jurisdiction. Id. at *8, 2006 WL 891187, at *2. Accordingly, the District Court directed BBI to submit, within 10 days of the Order, an affidavit setting forth the details of the fees and costs expended.

On May 5, 2006, BBI filed a timely notice of appeal of the portion of the District Court’s Order that vacated the California default judgment. On June 29, 2006, the New Jersey District Court stayed BBI’s application for fees and costs pending resolution of this appeal. The District Court had jurisdiction over the registration and enforcement of a foreign judgment pursuant to 28 U.S.C. § 1963 (2000), and we have jurisdiction over BBI’s *251appeal of a final order pursuant to 28 U.S.C. § 1291 (2000). The part of the District Court’s Order vacating the default judgment is “final” within the meaning of § 1291, despite the fact that the amount of attorneys’ fees remains unresolved. See Frangos v. Doering Equip. Corp., 860 F.2d 70, 72 (3d Cir. 1988) (citing Budinich v. Becton Dickinson, 486 U.S. 196, 199, 108 S.Ct. 1717, 100 L.Ed.2d 178 (1988)).

On May 17, 2006, the Defendants filed a notice of cross-appeal of the part of the Order that awarded fees and costs to BBI. Although neither party questioned the existence of appellate jurisdiction over this portion of the Order, we note that the District Court has not quantified the amount of attorneys’ fees, and “[i]t has long been the rule in this circuit that this court lacks jurisdiction to examine the merits of an attorneys’ fee award where the award has not been quantified.” Frangos, 860 F.2d at 72. We will vacate the award of fees and costs without considering its merits. The District Court granted the award under its “inherent power under Rule 60(b) to impose terms and conditions upon the opening of a judgment.” 2006 U.S. Dist. LEXIS 17207, at *8, 2006 WL 891187, at *2 (quoting Feliciano v. Reliant Tooling Co., 691 F.2d 653, 657 (3d Cir. 1982)). By vacating the Order opening the judgment, we vacate these conditions as well.

II.

Federal Rule of Civil Procedure 55(c) states that a court “may set aside a default judgment under Rule 60(b).” Rule 60(b) lists six reasons for which a “court may relieve a party or its legal representative from a final judgment, order, or proceeding,” including the catch-all provision in Rule 60(b)(6) that allows a court to relieve a party from a judgment for “any other reason that justifies relief’ aside from the more specific circumstances described in Rules 60(b)(l)-(5). We review grants or denials of relief under Rule 60(b), aside from those raised under Rule 60(b)(4),5 under an abuse of discretion standard. See Harris v. Martin, 834 F.2d 361, 364 (3d Cir. 1987).

The power of a court to invoke Rule 60(b) to vacate its own earlier judgment is unquestioned. As we discuss below, however, it is unclear whether a court has the power to invoke Rule 60(b) to vacate a judgment when the court in which the judgment is registered (the “registering court”) is different from the court that entered the judgment (the “rendering court”). Nothing in the text of Rule 55(c) or Rule 60(b) suggests that a registering court lacks the power to vacate the judgment of a different rendering court, but as we discuss below, several courts have suggested otherwise.

We decline to establish a categorical rule stating that registering courts lack the power to use Rule 60(b)(6) to vacate the judgments of rendering courts, but we emphasize that registering courts should exercise this power only under very limited circumstances. Even when a court is considering its own judgment, “extraordinary circumstances” must be present to justify the use of the Rule 60(b)(6) catch-all provision to vacate the judgment. See, e.g., Gonzalez v. Crosby, 545 U.S. 524, 535-36, 125 S.Ct. 2641, 162 L.Ed.2d 480 (2005) (citing Ackermann v. United States, 340 U.S. 193, 199, 71 S.Ct. 209, 95 L.Ed. 207 (1950)). When a court is considering whether to vacate another court’s judg*252ment under Rule 60(b)(6), these circumstances must be even more “extraordinary” because of the additional interest in comity among the federal district courts. We need not decide exactly how “extraordinary” a circumstance must be to justify the vacatur of another court’s judgment. The circumstances of the instant case would not even be extraordinary enough to justify a court’s decision to vacate its own judgment. It follows, a fortiori, that they are not extraordinary enough to justify vacating the judgment of another court.

A.

Several circuits have either held, or stated in dicta, that the power of a federal district court to set aside another district court’s judgment is limited.

In Indian Head National Bank v. Brunelle, 689 F.2d 245 (1st Cir. 1982), the First Circuit held that “a registration court errs in entertaining a Rule 60(b) motion that alleges neither a judgment void for lack of personal jurisdiction nor grounds that would support an independent equitable action.” Id. at 251-52. In Indian Head, the District Court for the Eastern District of Pennsylvania rendered a default judgment against the defendant, Conproeo Corporation. Id. at 247. The plaintiff registered the judgment in the District Court for the District of New Hampshire pursuant to 28 U.S.C. § 1963, and attempted to attach Conproco’s New Hampshire’s bank account. Id. Instead of directly appealing the judgment or moving in the rendering court to set it aside, Con-proco moved in the District of New Hampshire to set aside the judgment under Rule 60(b)(1) and Rule 60(b)(6), claiming that “the default judgment was due to inadvertence and neglect.” Id. at 247-48. The New Hampshire district court granted Conproco’s motion after finding that Con-proco had a valid defense and that the default judgment was not the fault of Con-proco or its counsel. Id. at 248. The First Circuit reversed, explaining that “there are indications that the drafters of the Rule intended to restrict motion practice under 60(b) to the court which rendered judgment,”6 and that deference to the rendering court promotes comity among federal courts, efficient judicial administration (in light of the rendering court’s familiarity with the issues), and simplified collection of judgments. Id. The First Circuit said that a registering court may grant Rule 60(b) relief from a different rendering court’s judgment in only two situations: first, when the request for relief is based on grounds that could also *253support an independent equitable action, id. at 249-50 & n. 8, and second, when a request for relief from a default judgment is made under Rule 60(b)(4), id. at 250-51. Conproco’s request for relief did not fit into these categories, so the First Circuit reversed the judgment of the district court granting the relief. Id. at 252.

The Seventh Circuit went further in Board of Trustees v. Elite Erectors, Inc., 212 F.3d 1031 (7th Cir. 2000), concluding that a rendering district court may not “modify or annul” the judgment of another district court under any provision of Rule 60(b), including Rule 60(b)(4). Id. at 1034. Elite Erectors involved a default judgment rendered in the District Court for the Eastern District of Virginia that the plaintiffs sought to enforce in the District Court for the Southern District of Indiana. Id. at 1033. The defendants moved in the Indiana district court to annul the judgment under Rule 60(b)(4) on the ground that the rendering court lacked personal jurisdiction over them. Id. The Seventh Circuit held that any request to annul or modify a judgment under Rule 60(b) “must be presented to the rendering court.” Id. It proceeded to explain that the parties against whom the Virginia district court entered judgment “are entitled to resist enforcement in Indiana if, but only if, the United States District Court for the Eastern District of Virginia lacked personal or subject-matter jurisdiction.” Id. at 1035. But the Seventh Circuit indicated that “resisting enforcement” of the judgment in Indiana was not the same as “formally annulling it under Rule 60(b)(4).”7 Id. at 1034. The difference is that annulling the judgment would deprive it of its effect in all forums, including the rendering court, whereas successfully resisting the judgment’s enforcement would deprive it of its effect only in the Indiana district court. Id.

Several other circuits have suggested in dicta that Rule 60(b) motions should generally be made before the rendering court, but they have not adopted a rigid requirement. In Covington Industries, Inc. v. Resintex A.G., 629 F.2d 730 (2d Cir. 1980), the Second Circuit held that a registering court could invoke Rule 60(b)(4) to vacate a rendering court’s judgment for lack of jurisdiction, but it suggested that registering courts should be more restrained when applying other sections of Rule 60(b). See 629 F.2d at 733 (“In the usual case, the court of rendition will be more familiar with the facts than the court of registration and perhaps more conversant with the applicable law.”). In Morris ex rel. Rector v. Peterson, 871 F.2d 948 (10th Cir. 1989), the Tenth Circuit cited Indian Head for “the general rule that a registration court ... usually defers on Rule 60(b) motions to the court rendering the judgment ...,” but then applied Indian Head’s exception for Rule 60(b)(4) motions. Peterson, 871 F.2d at 950-51 & n. 2 (citing Indian Head, 689 F.2d at 249, 251-52). In Harper Macleod Solicitors v. Keaty & Keaty, 260 F.3d 389 (5th Cir. 2001), the Fifth Circuit held that a registering court may invoke Rule 60(b)(4) to vacate a rendering court’s judgment but expressed doubt in dicta that a registering court could do the same with other sections of Rule 60(b). See id. at 395 (“judicial efficiency and comity among district courts often counsel a registering court to defer ruling on Rule 60(b) motions in favor of the rendering court....”). In FDIC v. Aaronian, 93 F.3d 636 (9th Cir. 1996), the Ninth Circuit said that “[rjegis-*254tering courts generally prefer litigants to bring motions for postjudgment relief in the rendering court,” but it concluded that it was proper for a registering court to entertain a challenge to a rendering court’s judgment on the ground that the judgment was unconstitutional and therefore void.8 Id. at 649.

We are persuaded by the reasoning of the First, Second, Fifth, Seventh, Ninth, and Tenth Circuits to the extent that they conclude that Rule 60(b) motions (other than motions under Rule 60(b)(4)) should generally be raised in the rendering court. Nonetheless, we decline to hold that registering courts lack the power in all situations to invoke Rule 60(b)(6) to set aside judgments.9 Rule 60(b)(6) exists so that courts may “vacate judgments whenever such action is appropriate to accomplish justice,” Klapprott v. United States, 335 U.S. 601, 614, 69 S.Ct. 384, 93 L.Ed. 266 (1949), in situations that are not addressed by the other five clauses of Rule 60(b). The drafters of Rule 60(b)(6) apparently recognized that a catch-all provision would be necessary, since it would be impossible to specify all of the scenarios in which justice might require vacatur of a judgment. Given the catch-all nature of Rule 60(b)(6), we do not think that it would be wise to adopt a rule that categorically forbids district courts from vacating the judgments of other district courts under this provision.10 Although the interest in comity 11 will usually make it inadvisable for a registering court to vacate a rendering court’s judgment, we cannot rule out the possibility that some set of facts will cause the injustice of enforcing a rendering court’s judgment to be so great as to outweigh the damage that setting aside the judgment would inflict upon comity.12 We need not decide here what facts would be sufficient, however, because it is clear that *255the facts of the instant case fall far short of what is necessary to justify vacatur of a rendering court’s judgment under Rule 60(b)(6).

B.

Although the text of Rule 60(b)(6) states simply that a court may grant relief from a final judgment for “any other reason that justifies relief,” courts have added a requirement that a party seeking Rule 60(b)(6) relief must demonstrate the existence of “extraordinary circumstances”13 that justify reopening the judgment. See, e.g., Crosby, 545 U.S. at 535-36, 125 S.Ct. 2641 (citing Ackermann, 340 U.S. at 199, 71 S.Ct. 209); Coltec Indus. v. Hobgood, 280 F.3d 262, 273 (3d Cir. 2002) (quoting In re Fine Paper Antitrust Litig., 840 F.2d 188, 194 (3d Cir. 1988)). This requirement exists in order to balance the broad language of Rule 60(b)(6), which allows courts to set aside judgments for “any” reason justifying relief, with the interest in the finality of judgments. See In re Fine Paper Antitrust Litig., 840 F.2d at 194-95; Mayberry v. Maroney, 558 F.2d 1159, 1163 (3d Cir. 1977). As we explained above, when a registering court considers the judgment of a different district court, the use of Rule 60(b)(6) to vacate another court’s judgment implicates an additional interest in comity, even if the judgment was a default judgment. If the circumstances of a case are not sufficiently “extraordinary” to outweigh the interest in the finality of judgments, then it follows that the circumstances cannot outweigh the interest in finality combined with the interest in comity.

We have explained that a showing of extraordinary circumstances involves a showing that without relief from the judgment, “an ‘extreme’ and ‘unexpected’ hardship will result.” Mayberry, 558 F.2d at 1163. This “hardship” requirement may sometimes be satisfied when the judgment “precluded an adjudication on the merits.” Boughner v. Sec’y of Health, Educ. & Welfare, 572 F.2d 976, 978 (3d Cir. 1978). But extraordinary circumstances rarely exist when a party seeks relief from a judgment that resulted from the party’s deliberate choices.14 See, e.g., Coltec, 280 F.3d at 274 (“[CJourts have not looked favorably on the entreaties of parties trying to escape the consequences of their own ‘counseled and knowledgeable’ decisions.”); see also Ackermann, 340 U.S. at 198-99, 71 S.Ct. 209 (petitioner could not show the existence of extraordinary circumstances when he voluntarily chose not to appeal due to the modest expenses that an appeal would require).

In the instant case, the New Jersey District Court did not mention the “extraordinary circumstances” requirement in its opinion. We acknowledge that our decision in Harad may have contributed to this error by sending confusing signals to the District Court regarding the standards for vacating a default judgment. In Har-*256ad, we reversed the district court’s decision not to vacate its own earlier default judgment. 839 F.2d at 985. We stated that “the decision to vacate a default judgment is left to the sound discretion of the trial court,” but that “[i]n exercising this discretion ... the court must consider whether vacating the default judgment will visit prejudice on the plaintiff, whether the defendant has a meritorious defense, and whether the default was the result of the defendant’s culpable conduct.” Id. at 982. Harad makes no reference to any clause in Rule 60(b), nor does it consider whether “extraordinary circumstances” were present. Thus, Harad may have created the erroneous impression that an exception to the “extraordinary circumstances” requirement exists when a district court is considering whether to vacate a default judgment, as opposed to a judgment on the merits.

A closer look at Harad reveals that its test was not intended to apply to Rule 60(b)(6). Harad involved a complaint filed by an attorney (Charles Harad) and one of his insurance companies (Home), against another of his insurance companies (Aet-na), seeking a declaratory judgment that Aetna had a duty to defend and indemnify Harad. 839 F.2d at 981. On December 23,1986, Harad and Home served the complaint on Aetna at its Hartford office rather than the Philadelphia office with which they had negotiated previously. Id. The Hartford office forwarded the complaint to the Philadelphia office, and Aetna entered an appearance with the District Court for the Eastern District of Pennsylvania on January 9, 1987, the same day that Harad and Home filed a request for a default judgment. The district court docketed the default judgment request on January 12 and the entry of appearance on January 14, and then granted the request for default judgment on January 14. Id. Aetna moved to vacate the default judgment, and the parties stipulated that the sole issue to be addressed on the motion was whether Aetna had established a meritorious defense to the plaintiffs’ action. Id. The district court then declined to vacate the default judgment, but we reversed, concluding that Aetna had a meritorious defense. Id. at 985. Although neither the district court’s opinion, see Harad v. Aetna Cas. & Sur. Co., No. 86-cv7266, 1987 WL 12290 (E.D.Pa. June 9, 1987), nor our opinion cited any Federal Rule of Civil Procedure, we see at least two reasons for construing Harad as a grant of relief under Rule 60(b)(1), which allows relief from a judgment on the basis of “mistake, inadvertence, surprise, or excusable neglect,” rather than Rule 60(b)(6). First, the facts of the case, with the complaint sent to the wrong office and the apparent delay in docketing the entry of appearance, suggest that this was a matter of “mistake, inadvertence, surprise, or excusable neglect.” Second, Harad obtains its three-part test directly from United States v. $55,518.05 in U.S. Currency, 728 F.2d 192 (3d Cir. 1984), a case that addresses Rule 60(b)(1). See Harad, 839 F.2d at 982 (citing $55,518.05 in U.S. Currency, 728 F.2d at 195). In $55,518.05 in U.S. Currency, we said:

We require the district court to consider the following factors in exercising its discretion in granting or denying a motion to set aside a default under Rule 55(c) or a default judgment under Rule 60(b)(1): (1) whether the plaintiff will be prejudiced; (2) whether the defendant has a meritorious defense; (3) whether the default was the result of the defendant’s culpable conduct.

$55,518.05 in U.S. Currency, 728 F.2d at *257195 (emphasis added).15 Accordingly, we conclude that the District Court erred by relying solely on the three factors listed in Harad to grant relief under Rule 60(b)(6).

Appellees assert that Emcasco Insurance Co. v. Sambrick, 834 F.2d 71 (3d Cir. 1987), provides the appropriate factors for deciding whether to set aside a default judgment. In Emcasco, we applied the same three factors that we used in Harad, plus a fourth — “the effectiveness of alternative sanctions.” Id. at 73. Emcasco does not cite Rule 60(b) in its review of the district court’s decision not to vacate a default judgment. As with Harad, however, we construe Emcasco as describing a standard for vacating a default judgment under Rule 60(b)(1), not Rule 60(b)(6). First, we think that the default judgment in Emcasco could be fairly described as the product of “mistake, inadvertence, surprise, or excusable neglect.”16 The Em-casco district court decided on January 14, 1987 that it would enter a default judgment against the defendant unless he filed an answer by January 16. 834 F.2d at 75. But the defendant and his counsel were never informed of this decision and instead were erroneously told that they could avert a default judgment merely by entering an appearance, which they did on January 15. Id. Given these facts, Emcasco was in effect a Rule 60(b)(1) case. Second, most of the cases that Emcasco cites in support of its four-part test are applications of Rule 60(b)(1). See 834 F.2d at 73-74.17 Thus, Emcasco does not provide a *258test that is applicable to a Rule 60(b)(6) motion.

We acknowledge that default judgments are generally disfavored in our circuit. See $55,518.05 in U.S. Currency, 728 F.2d at 194-95 (“[Tjhis court does not favor entry of defaults or default judgments. We require doubtful cases to be resolved in favor of the party moving to set aside the default judgment ‘so that cases may be decided on their merits.’ ” (quoting Tozer v. Charles A. Krause Milling Co., 189 F.2d 242, 244 (3d Cir. 1951))). But we cannot apply this presumption against default judgments if doing so would be inconsistent with the Federal Rules of Civil Procedure or our case law interpreting these Rules. Because “extraordinary circumstances” are essential for a grant of Rule 60(b)(6) relief, a case does not become “doubtful” when the district court has made no attempt to show that such circumstances exist.

Finally, our own review of the record does not suggest that the circumstances of this case are “extraordinary” as we have defined this term in our case law. The Defendants acknowledge that their decision not to contest the California judgment was the result of a deliberate choice. Thus, the default judgment cannot be said to have created an “unexpected hardship.” Cf. Boughner, 572 F.2d at 978. Since nothing else in the record appears to qualify as an “extraordinary circumstance” sufficient to justify setting aside any judgment, let alone one entered by a different court, we will vacate the District Court’s Order setting aside the default judgment.

III.

We next confront a question that parallels the one we addressed about Rule 60(b)(6): does a federal district court have the power to consider a motion to vacate another district court’s judgment under Rule 60(b)(4) on the ground that the latter court lacked personal jurisdiction over the defendant? Neither party has contended that the District Court lacks this power. Nevertheless, we will explain why we hold that this power exists.

Rule 60(b)(4) allows a court to relieve a party from a final judgment if “the judgment is void.” A judgment is void within the meaning of Rule 60(b)(4) if the court that rendered it lacked personal jurisdiction over the defendant. See Marshall v. Bd. of Educ., 575 F.2d 417, 422 (3d Cir. 1978). Although there is no question that a court may grant a Rule 60(b)(4) motion to vacate one of its own judgments, we have not explicitly ruled on whether a court has the power to grant a Rule 60(b)(4) motion to vacate another court’s judgment for lack of personal jurisdiction.18

*259Five circuits have stated either in holdings or dicta that a registering court has the power to hear a Rule 60(b)(4) motion. Three of these circuits explicitly held that a registering court may vacate a default judgment of a rendering court when the rendering court lacked personal jurisdiction over the defendant. Covington, 629 F.2d at 732-34 (Second Circuit); Harper Macleod, 260 F.3d at 394-95 (Fifth Circuit); Peterson, 871 F.2d at 951 & n. 2 (Tenth Circuit). The First Circuit endorsed the view of these circuits in dicta, recognizing that Rule 60(b)(4) challenges to default judgments are one of the two exceptions to the general rule that Rule 60(b) motions must be addressed to the rendering court. See Indian Head, 689 F.2d at 250-51. The Ninth Circuit endorsed a registering court’s power to vacate a rendering court’s judgment under Rule 60(b)(4), but its holding addressed a Rule 60(b)(4) motion based on a challenge to the constitutionality of the rendering court’s judgment rather than the rendering court’s personal jurisdiction. 93 F.3d at 639-40. We note that the Eastern District of Pennsylvania has produced a thorough and well-reasoned opinion in On Track Transportation, Inc. v. Lakeside Warehouse & Trucking, Inc., 245 F.R.D. 213 (E.D.Pa. 2007), in which it held that a registering court has the power to decide a Rule 60(b)(4) motion challenging a rendering court’s default judgment for lack of subject matter jurisdiction. Id. at 214-23.

The Seventh Circuit is an outlier on this issue. As we discussed above, the Seventh Circuit concluded in Elite Erectors that the registering court “was free to disregard the judgment, without formally annulling it under Rule 60(b)(4), if the rendering court lacked [personal or subject-matter] jurisdiction.” 212 F.3d at 1034-35. Because the registering court cannot annul the judgment under Rule 60(b)(4), the most it can do is decline to enforce the judgment — while leaving the judgment in place so that other jurisdictions can enforce it. See id. As the Eastern District of Pennsylvania noted in On Track, this is an “impracticable” solution. See 245 F.R.D. at 219-20. If the rendering court merely issued a default judgment, then it never actually litigated the issue of personal jurisdiction. In contrast, the registering court will have actually litigated the issue. Thus, the registering court’s judgment will have preclusive effect, and a defendant “could then move the rendering court to vacate the judgment on the basis that it lacked jurisdiction, using the decision of the registering court offensively.” Id. This would lead to the same result as vacating the judgment altogether, except with one extra step, along with the expenditure of time and resources that this step would entail. Id. at 220. We see no reason to break with the majority of circuits and embrace the cumbersome process endorsed by the Seventh Circuit.

Finally, we do not think a registering court seriously threatens the interest in comity when it vacates a rendering court’s default judgment under Rule 60(b)(4) for lack of personal jurisdiction. If the rendering court did not have personal jurisdiction, then the judgment was not merely erroneous; it never should have been entered in the first place. Moreover, when the rendering court enters a default judgment based on nothing but one party’s failure to appear, there is no risk that the courts will reach opposite conclusions on personal jurisdiction because the rendering court was not required to address the issue. Finally, as the On Track court noted, comity “must be balanced against the longstanding principle that ‘[a] defendant is always free to ignore the judicial proceedings, risk a default judgment, and then challenge that judgment on jurisdictional grounds in a collateral proceeding.’ ” *260245 F.R.D. 213, 221 (citing Ins. Corp. of Ir., Ltd. v. Compagnie des Bauxites de Guinee, 456 U.S. 694, 706, 102 S.Ct. 2099, 72 L.Ed.2d 492 (1982)). The latter principle would be undermined if we held that Rule 60(b)(4) actions could only be brought in the rendering court that allegedly lacked personal jurisdiction over the defendant.

IV.

We now reach the question that the parties originally asked the District Court to address: whether the California district court possessed personal jurisdiction over the Defendants. We have the power to resolve this issue ourselves, rather than remand it to the District Court, because we conduct de novo review of jurisdictional issues raised under Rule 60(b)(4). See Page v. Schweiker, 786 F.2d 150, 152 (3d Cir. 1986). Both parties have informed us in their briefs and at oral argument that they would prefer that we resolve the issue rather than remand it. Unfortunately, we are unable to do so, because we find that the existence of personal jurisdiction turns on disputed questions of fact that should be resolved by a factfinder. Therefore, we remand the matter to the District Court, with the hope that the District Court will resolve this matter as promptly and as inexpensively for the parties as possible. Below we provide an overview of the personal jurisdiction issues in this case, pointing out what we believe are the key factual disputes.

A. Personal Jurisdiction in California

“In a diversity action, [a California federal district court] may exercise personal jurisdiction over a non-resident defendant if jurisdiction is proper under California’s long-arm statute and if that exercise of jurisdiction accords with federal constitutional due process principles.” Fireman’s Fund Ins. Co. v. Nat’l Bank of Coops., 103 F.3d 888, 893 (9th Cir. 1996). California’s long-arm statute provides simply that the court may exercise personal jurisdiction “over a non-resident defendant on any basis not inconsistent with the California or federal Constitution.” Id. (citing Cal.Code Civ. Peoc. § 410.10). “The statutory and constitutional requirements therefore merge into a single due process test,” which requires that the defendant “have certain minimum contacts with [the forum] such that the maintenance of the suit does not offend traditional notions of fair play and substantial justice.” Id. (citing Int’l Shoe Co. v. Washington, 326 U.S. 310, 316, 66 S.Ct. 154, 90 L.Ed. 95 (1945) (internal quotation omitted)). As both parties acknowledge, we determine whether this requirement is satisfied by applying a three-part test. The three-part test, as summarized in Panavision International v. Toeppen, 141 F.3d 1316 (9th Cir. 1998), is as follows:

(1) The nonresident defendant must do some act or consummate some transaction with the forum or perform some act by which he purposefully avails himself of the privilege of conducting activities in the forum, thereby invoking the benefits and protections of its laws; (2) the claim must be one which arises out of or results from the defendant’s forum-related activities; and (3) exercise of jurisdiction must be reasonable.

141 F.3d at 1320 (citation omitted). BBI contends that all three of these requirements were satisfied as a matter of law.19

*261B. Purposeful Availment

BBI asserts that White has purposefully-availed herself of the privilege of conducting activities in California and thereby has invoked the benefits of California law. BBI bases its argument on the Settlement Agreement as well as the Defendants’ alleged trademark infringement.

1. Purposeful Availment Through The Settlement Agreement

The mere existence of a contract is insufficient to establish minimum contacts. Burger King Corp. v. Rudzewicz, 471 U.S. 462, 478, 105 S.Ct. 2174, 85 L.Ed.2d 528 (1985) (“If the question is whether an individual’s contract with an out-of-state party alone can automatically establish sufficient minimum contacts in the other party’s home forum, we believe the answer clearly is that it cannot.”). But a contract is typically an intermediate step between past negotiations and future transactions, and Burger King instructs that “it is these factors — prior negotiations and contemplated future consequences, along with the terms of the contract and the parties’ actual course of dealing — that must be evaluated in determining whether the defendant purposefully established minimum contacts with the forum.” Id. at 479, 105 S.Ct. 2174. BBI argues that several aspects of the Settlement Agreement’s terms, as well as prior negotiations and contemplated future consequences, create minimum contacts.

First, BBI points out that the Agreement contains a choice-of-law clause stating that it “will be governed by and construed under the laws of the State of California.” Br. of Appellant 39. As Burger King points out, a choice-of-law provision “standing alone would be insufficient to confer jurisdiction,” but combined with other factors, it may reinforce a party’s “deliberate affiliation with the forum State and the reasonable foreseeability of possible litigation there.” 471 U.S. at 482, 105 S.Ct. 2174. We must therefore consider other factors before we can decide how much weight to afford to this provision.

Second, BBI points out that the Defendants negotiated the Settlement Agreement by telephone and mail with California-based BBI and its attorneys. Br. of Appellant 40. Interpreting California law, the Ninth Circuit has said that, ordinarily, use of the mails and telephone “simply do not qualify as purposeful activity invoking the benefits and protection of the [forum] state.” Roth v. Garcia Marquez, 942 F.2d 617, 622 (9th Cir. 1991) (quoting Thos. P. Gonzalez Corp. v. Consejo Nacional De Produccion De Costa Rica, 614 F.2d 1247, 1254 (9th Cir. 1980)). Moreover, an important distinction between the negotiations in Burger King and those in the instant case is that the Burger King defendant actively sought contract negotiations with a company based in the forum state, whereas White did not reach out to anyone in California until BBI threatened litigation. See Burger King, 471 U.S. at 479, 105 S.Ct. 2174 (“Eschewing the option of operating an independent local enterprise, Rudzewicz deliberately ‘[reached] out beyond’ Michigan and negotiated with a Florida corporation for the purchase of a long-term franchise and the manifold benefits that would derive from affiliation with a nationwide organization.”).

Third, BBI asserts that the Defendants established minimum contacts “by accepting BBI’s settlement payment which was wired from a California bank to defendants.” Br. of Appellant 41. To underscore the importance of this payment, BBI cites a case from the Southern District of Mississippi, Medical Assurance Co. of Mississippi v. Jackson, 864 F.Supp. 576 (S.D.Miss. 1994). Jackson involved a suit brought by a Mississippi-based medical *262malpractice insurance company against two Alabama residents: Moore, the victim of a botched operation in Mississippi that created a need for additional surgery; and Jackson, Moore’s attorney. Id. at 577. Jackson threatened to sue the insurance company on Moore’s behalf and demanded a $1,100,000 settlement. Id. After “a series of letters and telephone calls,” Moore agreed to a settlement of $56,250 from the insurance company in exchange for an absolute release. Id. Jackson and Moore accepted and negotiated the $56,250 check that the insurance company tendered, but they refused to execute the absolute release that the company had also sent them. Id. They returned an altered version of the release form instead. Id. at 578. The insurance company sued Jackson and Moore in a Mississippi federal district court, alleging breach of the settlement agreement. Id. at 578. Jackson and Moore claimed that the court lacked personal jurisdiction over them because the insurance company had sent the payment and the release documents to Alabama. Id. The court rejected their claim, pointing out that “the check was sent from Mississippi and was ultimately paid by a Mississippi bank,” that “the release, which is central to this action” was sent from Mississippi and was to be returned to Mississippi, and that “defendants did, in fact, return a [modified] release document” to Mississippi. Id. Moreover, the court pointed out that Jackson initiated the transaction by threatening to sue the Mississippi-based insurance company. Id. at 579.

Despite some superficial similarities, Jackson is distinguishable. Although White and her attorney, like Moore and Jackson, accepted payment from a bank located in the forum state in exchange for accepting a settlement agreement, they did not initiate the transaction by threatening to sue a California company.20 Also, whereas Moore had traveled to Mississippi for surgery performed by a Mississippi doctor, 864 F.Supp. at 577, White had not directed any activities toward California before BBI contacted her. Thus, even if we regard Jackson as persuasive authority, it does not compel a finding of purposeful availment.

Fourth, BBI argues that the parties contemplated future consequences. Among other things, the Agreement states that the Defendants “hereby irrevocably appoint BBI as their respective lawful attorney-in-fact with authority to file any document in the name of and on behalf of [the Defendants] for the purpose of taking any of the actions required by [Section 4 of the Agreement] upon the event of a default.” We agree that the attorney-in-fact provision suggests that the parties contemplated future consequences to some extent. Still, a large gulf exists between the future contacts contemplated by this provision and the extensive future contacts contemplated by the Burger King contract. Whereas the Burger King defendant “entered into a carefully structured 20-year relationship that envisioned continuing and wide-reaching contacts with Burger King in Florida,” id. at 480, 105 S.Ct. 2174, the Settlement Agreement merely describes what will happen in the event of default.

In sum, we question whether any of the factors that BBI cites (the choice of law clause, the mail and telephone negotiations, the acceptance of the settlement check, and the attorney-in-fact provision) would individually support purposeful availment. The combined force of these factors, however, may be sufficient. On the other hand, White has asserted that *263the original agreement contained a clause consenting to personal jurisdiction in California that the parties removed at White’s insistence. If true, this suggests that we should hesitate to read the terms of the contract, especially the choice-of-law clause, as indicative of the “reasonable foreseeability of possible litigation” in California court.21 See 471 U.S. at 482, 105 S.Ct. 2174. BBI contests this fact, and we do not have evidence of it in the record other than White’s affidavit and brief. Therefore, on remand the District Court may wish to investigate White’s claim regarding the contract negotiations, since it could shed light on the extent to which the parties contemplated an ongoing relationship with California, including the possibility of litigation there.22

2. Purposeful Availment Through Trademark Infringement

BBI contends that the Defendants’ alleged trademark infringement created personal jurisdiction in California because the Defendants purposefully directed their activities toward that forum. Br. of Appellant 42. Relying on the “effects test” of Calder v. Jones, 465 U.S. 783, 104 S.Ct. 1482, 79 L.Ed.2d 804 (1984), BBI says that purposeful availment exists if the defendant “(1) committed an intentional act, (2) expressly aimed at the forum state, (3) causing harm that the defendant knows is likely to be suffered in the forum state.” Dole Food Co. v. Watts, 303 F.3d 1104, 1111 (9th Cir. 2002) (summarizing CaldePs “effects test”).

BBI contends that trademark infringement “is treated as an intentional tort for jurisdictional purposes.” The first case that BBI cites for this proposition is Sebastian Int’l, Inc. v. Russolillo, No. 00-03476 CM, 2000 U.S. Dist. LEXIS 21510 (C.D.Cal. Aug. 25, 2000). In Russolillo, the court held that the defendant’s trademark infringement through the sale of counterfeit goods was “intentional” because the plaintiff had written a letter to put the defendant on notice of the infringement. 2000 U.S. Dist. LEXIS 21510, at *13. BBI claims that Russolillo is analogous to the instant case, because BBI informed the Defendants of their infringement in 2003, prior to the settlement, and informed them of their continuing infringement on February 1, 2005. The second case that BBI cites is Panavision. In Panavision, the court stated that although the defendant’s registration of the plaintiffs trademark as a domain name was not sufficient to establish jurisdiction in the forum state, the defendant’s “scheme to register [the plaintiffs] trademarks as his domain names for the purpose of extorting money” from the plaintiff was an intentional act directed at the forum state.23 141 *264F.3d at 1322. The instant case is distinguishable from both Russolillo and Pana-vision. In Russolillo, the court indicated that it was the explicit warning about trademark infringement, not the act of counterfeiting itself, that rendered the defendant’s trademark infringement “intentional.” 2000 U.S. Dist. LEXIS 21510, at *13. Although BBI also sent a warning to the Defendants, the record suggests that the Defendants acted in good faith to remedy the alleged infringement. If this is true, then any ongoing “infringement” was not the result of intentional acts on the Defendants’ part — unless we agree with BBI’s claim that surrendering control of the telephone number was absolutely necessary to prevent further infringement. Panavision supports the Defendants, because it makes clear that the mere use of another party’s trademark (i.e., registration of domain names and creation of web sites) does not constitute an intentional act aimed at the forum state for jurisdictional purposes. 141 F.3d at 1322. It is possible, however, that the District Court’s findings of fact will reveal more about whether White’s behavior can be construed as “intentional.”

BBI also contends that the Defendants’ behavior was “expressly aimed at the forum state.” We do not find it immediately obvious why the failure to change New Jersey telephone listings and to surrender a New Jersey telephone number is behavior “expressly aimed” at California. But BBI relies on language in Bancroft & Masters v. Augusta National, 223 F.3d 1082 (9th Cir. 2000), which states that “the [‘expressly aiming’] requirement is satisfied when the defendant is alleged to have engaged in wrongful conduct targeted at a plaintiff whom the defendant knows to be a resident of the forum state.” 223 F.3d at 1086. Even though the Defendants’ alleged infringement was limited to activities in New Jersey, Bancroft implies that because it was targeted at franchisees of BBI, a known “resident” of California, it was therefore expressly aimed at California.24 But Panavision suggests a contrary result. Given the holding in Panavision that the use of a company’s trademarks in domain names and web sites is not “expressly aimed” at the forum in which the company has its principal place of business, 141 F.3d at 1322, we question whether listings in New Jersey telephone directories could be characterized as actions expressly aimed at California.

Finally, BBI contends that the “brunt of the harm” to its reputation and profits is felt in California because California is its principal place of business.25 BBI argues *265that just because BBI is a California corporation, a jurisdictionally sufficient amount of economic harm is felt in California. See Harris Rutsky & Co. Ins. Servs. v. Bell & Clements, Ltd., 328 F.3d 1122, 1131 (9th Cir. 2003) (for “a California corporation whose principal place of business is in California, ... the brunt of the harm [is] felt in California.”). We cannot resolve this issue without more evidence. BBI’s California complaint suggests that the Defendants’ alleged infringement is causing economic harm to BBI as a whole. But White claimed that as of March 9, 2005, her company had not received a single call from anyone confused by the erroneous listing, which would suggest that little economic harm has been inflicted upon the New Jersey franchisees, let alone the national entity. Without further factfinding, it is impossible to decide which party is correct about the economic effects felt by BBI in California.

C. Does The Cause of Action Arise Out of Forum-Related Activities?

The next prong of the three-part test for specific jurisdiction is whether the cause of action arises out of forum-related activities. Although this is a distinct requirement, in this case the analysis is identical to the “purposeful availment” analysis. BBI has not asserted that White or her company has any connection to California aside from those associated with the causes of action for breach of contract and trademark infringement. Thus, if the first prong is satisfied, then the second prong is satisfied as well.

D. Fair Play and Substantial Justice, a.k.a. Reasonableness

If the District Court finds that the first two prongs are satisfied, it must consider whether “minimum requirements inherent in the concept of ‘fair play and substantial justice’ may defeat the reasonableness of jurisdiction even if the defendant has purposefully engaged in forum activities.” Burger King, 471 U.S. at 477-78, 105 S.Ct. 2174. Although there is a “strong presumption of reasonableness” if purposeful availment exists, Dole Food, 303 F.3d at 1117, courts must still consider reasonableness as a separate factor. This is a heavily fact-based inquiry. California has identified seven factors that affect reasonableness, none of which are determinative:

(1) the extent of the defendants’ purposeful interjection into the forum state’s affairs; (2) the burden on the defendant of defending in the forum; (3) the extent of conflict with the sovereignty of the defendants’ state; (4) the forum state’s interest in adjudicating the dispute; (5) the most efficient judicial resolution of the controversy; (6) the importance of the forum to the plaintiffs interest in convenient and effective relief; and (7) the existence of an alternative forum.

Harris Rutsky, 328 F.3d at 1132 (citing Core-Vent Corp. v. Nobel Indus. AB, 11 F.3d 1482, 1487-88 (9th Cir. 1993)). Most of these factors cannot be resolved without factfinding from the District Court. For example, the District Court might wish to consider the second factor in light of the statements throughout White’s briefs that it is extremely burdensome for a small business in New Jersey with no California connections (aside from, perhaps, those created through its dealings with BBI) to litigate in California. To decide the fifth factor, the District Court should “focus on *266the location of the evidence and witnesses.” Harris Rutsky, 328 F.3d at 1133. If most of the evidence and witnesses are located in New Jersey, where the alleged infringement occurred, this factor weighs against personal jurisdiction in California.

V.

We will vacate the April 5, 2006 Order of the District Court for the District of New Jersey, including the assessment of fees and costs. We remand this case so that the District Court may decide the question that the parties asked it to decide over two years ago: whether to set aside the California default judgment for lack of personal jurisdiction.

concurring in part and dissenting in part.

This case arises out of the context of a 28 U.S.C. § 1963 registration proceeding, and more fundamentally, involves a collateral attack to a foreign default judgment. It is well-settled that the grounds for obtaining collateral relief are extremely limited. The majority, however, holds that registering courts have the power to vacate a foreign judgment rendered by their sister federal courts pursuant to Federal Rule of Civil Procedure 60(b)(6). In doing so, it becomes the first court in the nation to reach this conclusion. Because I think this holding potentially creates a circuit split where none existed before, and because I question whether' such a broad ruling is necessary under the particular circumstances of this case, I respectfully dissent.

Under the circumstances before us, one may only seek to collaterally vacate a default judgment obtained in another jurisdiction based on a challenge to the underlying validity of the judgment. Sheet Metal Workers’ Nat’l Pension Fund v. Elite Erectors, Inc., 212 F.3d 1031, 1034 (7th Cir. 2000) (“[a] party that simply refuses to appear may contend in a later case that the first tribunal lacked jurisdiction — though jurisdiction is the only issue thus preserved”) (emphasis in original); Yale v. Nat’l Indem. Co., 602 F.2d 642, 644 (4th Cir. 1979) (“only void judgments are subject to collateral attack, and [ ] a void judgment is only one that is rendered by a court lacking jurisdiction over the default or over the subject matter”); see also Ins. Corp. of Ireland, Ltd. v. Compagnie des Bauxites de Guinee, 456 U.S. 694, 706, 102 S.Ct. 2099, 72 L.Ed.2d 492 (1982) (“[a] defendant is always free to ignore the judicial proceedings, risk a default judgment, and then challenge that judgment on jurisdictional grounds in a collateral proceeding”). From this, it naturally follows that registering courts may entertain motions to vacate registered foreign judgments pursuant to Rule 60(b)(4), on the basis that the rendering court lacked jurisdiction. See 12 James Wm. Moore et al., Moore’s Federal Practice § 60.60[3][c] (3d ed. 1997); 30 Am.Jur.2d Executions and Enforcement of Judgments § 813 (“the court in which the judgment is registered may determine whether the court which entered the judgment had jurisdiction over the subject matter and parties”). This is certainly the view that has been expressly embraced by nearly all26 of our sister circuits to have squarely confronted the question, and I thus concur with the majority’s opinion *267insofar as it adopts this relatively uncontroversial position as Third Circuit law.27

However, the opinion does not stop there, but goes further to hold that registering courts may also consider Rule 60(b)(6) motions to vacate foreign judgments under certain “extraordinary circumstances.” Here, I must part ways with my majority colleagues. While it is true that the precise contours of the registering court’s authority to grant Rule 60(b) relief remain unsettled, I am nevertheless unaware of any court considering the question — before today — that has affirmatively declared such authority to exist for any Rule 60(b) subsection outside of the (b)(4) context.28 See 11 Charles A. Wright, Arthur R. Miller & Mary Kay Kane, FedeRal PRactice & Procedure § 2787, at 36 (2d ed. 1995) (“the courts have not squarely decided to what extent, if at all, the court in which a judgment is registered can give relief from the judgment under Rule 60(b)”); Indian Head Nat’l Bank of Nashua v. Brunelle, 689 F.2d 245, 251 (1st Cir. 1982) (noting neither the parties nor the court itself could find a single case “where a court of registration was willing to entertain directly a Rule 60(b) motion other than one attacking a default judgment for lack of personal jurisdiction”). Indeed, the only court directly confronting *268the issue expressly held to the contrary: “[A] registration court errs in entertaining a Rule 60(b) motion that alleges neither a judgment void for lack of personal jurisdiction nor grounds that would support an independent equitable action.” Brunelle, 689 F.2d at 251-52.

This is not surprising, in light of the general rule that notwithstanding the rule’s silence on the topic, applications for Rule 60(b) relief must typically be made in the court rendering the judgment. See, e.g., 12 James Wm. Moore et al., MooRe’s Fedeeal PRACTICE § 60.60[1] (3d ed. 1997) (although the rule itself does not expressly so provide, “it is clear that the drafters of the rule contemplated that the motion ... would always be brought ‘in the court and in the action in which the judgment was rendered’ ”) (emphasis added); 11 Charles A. Wright, Arthur R. Miller & Mary Kay Kane, FedeRal Practice & PROcedure § 2865, at 377 (Rule 60(b) motions generally made to rendering court). While this does not per se preclude one from challenging the underlying judgment upon registration, registering courts most often deny such challenges without prejudice, referring the parties to litigate the Rule 60(b) issue in the rendering court. See 11 Federal Practice & Procedure § 2865, at 378 (“it is appropriate for the court in the district of registration to decline to pass on the motion for [Rule 60(b) ] relief and to require the moving party to proceed in the court that gave judgment”); e.g., United States v. Fluor Corp., 436 F.2d 383, 384-85 (2d Cir. 1970) (affirming registration court’s denial of Rule 60(b)(5) motion without prejudice). And rightly so. With these well-settled legal principles in mind, I think the First Circuit’s holding in Brunette best vindicates the significant comity interests implicated in the registration context while adequately preserving collateral avenues of relief for litigants.

Furthermore, even if one agrees with the substance of the majority’s legal conclusion — one that is at odds with the Brunette decision — -that registering courts have the power to vacate foreign judgments under certain unspecified “extraordinary circumstances” pursuant to Rule 60(b)(6), the question remains whether this is the appropriate case in which to make such a proclamation. This is especially so when, notwithstanding its articulation of a new rule expanding the bases for which collateral relief may be granted in registration proceedings, the majority nevertheless acknowledges that no extraordinary circumstances exist in this case, in any event, so as to warrant any such relief. It would seem then, that a satisfactory resolution could have been had here without advancing the Rule 60(b)(6) issue; we should have accordingly done simply this. For me, this case begins and ends with the District Court’s vacatur of the default judgment insofar as its decision was based in any part on its perception of the merits of the underlying California action.29 It is axiomatic that when a party purposefully fails to appear in an action on the basis of the belief that the foreign tribunal lacks personal jurisdiction, she waives the right *269to later contest the underlying merits of that action in a collateral proceeding. E.g., Elite Erectors, Inc., 212 F.3d at 1034; Hazen Research, Inc. v. Omega Minerals, Inc., 497 F.2d 151, 154 (5th Cir. 1974) (where “the defendant makes no appearance and the judgment goes by default, the defendant may defeat subsequent enforcement in another forum by demonstrating that the judgment issued from a court lacking personal jurisdiction ... [but] should the attack fail, the default judgment becomes no less final and determinative on the merits of the controversy than a decree entered after full trial”). Thus, when the District Court considered the merits of the California action below to opine that defendant “possesses an extremely meritorious defense” warranted vacatur, App. vol. 1 at 7, it missed the fundamental distinction between direct and collateral attacks. 18 Moore’s Federal Practioe § 130.35[1] (“registering court cannot look into the substance of the judgment to reexamine its validity”). As the Restatement cogently explains:

[I]t is inappropriate to consider the merits of an attack on a judgment when that attack is made in the course of a subsequent action in which the judgment is relied on as a basis of claim or defense. To consider the merits of the attack in such a context is to contravene the general principle that relief from a judgment should be sought in the court that rendered the judgment.

RESTATEMENT (SECOND) OF JüDGMENTS § 80 cmt. a ¶ 2 (1982).

Accordingly, although I concur with the majority’s ultimate disposition of the case, I respectfully dissent from its pronouncements on the general availability of Rule 60(b)(6) relief. Under the particular circumstances here, I would hold that a registration court errs when it vacates a foreign judgment on grounds other than those of voidness or which would otherwise support an independent equitable action. The Full Faith and Credit Clause and the weighty interests of comity demand nothing less.