*995MEMORANDUM *

Eva Van Schijndel1 appeals the district court’s dismissal of her claims against The Boeing Company and Goodrich Corporation (collectively, “Boeing”) on forum non conveniens grounds. For several reasons, we conclude that the district court’s assessment of the relevant factors was not reasonable and, therefore, that the dismissal constituted an abuse of discretion. Ravelo Monegro v. Rosa, 211 F.3d 509, 511 (9th Cir. 2000).

First, the decision misstates the plaintiffs’ theory as being “that a fully loaded, fully fueled 747 aircraft, moving down a closed, under-construction runway at take-off speed, should be able to withstand collision with construction cranes, bulldozers, and the like.” This characterization differs from the district court’s earlier and correct understanding of the action as premised on the malfunctioning of the allegedly flawed design and manufacture of the plane’s emergency and evacuation equipment, not the “integrity of the aircraft” itself on impact. The difference between these theories is material with respect to several key inquiries pertinent to the forum non conveniens question, including which evidence and witnesses are material and whether the products liability case would be more conveniently heard in the same location as the case against Singapore Airlines.

Second, unlike the initial order denying Boeing’s original motion to dismiss, which names and rejects Singapore as the single alternative forum, the order upon reconsideration names three alternative forums: Singapore, Canada, and Taiwan. As a result, the court failed to balance the competing interests fairly by comparing the domestic forum to a particular foreign forum, and it is unclear which alternative forum the court ultimately found to be both adequate and more convenient than the domestic forum chosen by plaintiffs. Compare Piper Aircraft Co. v. Reyno, 454 U.S. 235, 238-41, 257-61, 102 S.Ct. 252, 70 L.Ed.2d 419 (1981) (comparing domestic forum with a single forum, Great Britain); Lueck, 236 F.3d at 1143-48 (comparing domestic forum with a single foreign forum, New Zealand); and Cheng v. Boeing Co., 708 F.2d 1406, 1410-12 (9th Cir. 1983) (comparing domestic forum with a single foreign forum, Taiwan) with, ER 88-92 (grouping together multiple foreign fora and comparing with domestic forum). This defect is of great significance on the facts of the case, as the crash took place in Taiwan, the airline is headquartered in Singapore, and the relevant witnesses appear to reside in Singapore, Taiwan, the United States, and elsewhere. Additionally, at the time of this appeal, the only remaining plaintiffs are from the Netherlands. Consequently, unlike in Piper Aircraft, Lueck, and Cheng, it is far from evident that there is any single alternative forum more convenient than the United States.

Third, the district court's statement in its dismissal order that a joint trial of Boeing and Singapore Airlines could risk prejudice to each defendant does not comport with the finding, in the same order, that the existence of various foreign fora, in each of which all claims might be re*996solved in a single action, supported dismissal. Compare ER 86 and ER 91. Although the district court may have had reasons for reaching this seemingly contradictory conclusion, those reasons do not appear in the record.

These three errors in the district court’s overall approach to its forum non conveniens analysis so affected the court’s balance of the private and public interest factors as to render that balance unreasonable.

Therefore, we REVERSE the district court and REMAND for further proceedings consistent with this disposition.

REVERSED AND REMANDED. .

The Supreme Court observed in Piper Aircraft Co. v. Reyno, 454 U.S. 235, 102 S.Ct. 252, 70 L.Ed.2d 419 (1981), “[t]he forum non conveniens determination is committed to the sound discretion of the trial court.... In examining the District Court’s analysis of the public and private interests, however, the Court of Appeals seems to have lost sight of this rule, and substituted its own judgment for that of the District Court.” Id. at 257, 102 S.Ct. 252. Much the same can be said here. Rather than deferring, the majority nitpicks the district court’s order, finding fault where none exists. Specifically, my colleagues hold that the district court committed three errors, but they are thrice mistaken.

The majority’s first claim of error is that the district court “misstates the plaintiffs’ theory [of liability]” and is confused about whether this case is about the “integrity of the aircraft” or “the malfunctioning of the allegedly flawed design and manufacture of the plane’s emergency and evacuation equipment.” Maj. mem. at 3 (internal quotation marks omitted). But as the majority acknowledges, id., the district court transcript shows that the court understood very well that appellants’ claims related to the aircraft’s safety features, see, e.g., ER 57 (“Indeed, what I know and what I’m growing to know more and more about in terms of the accident itself might make me raise my eyebrows about claims of product liability, but I do know that there are at least issues pertaining to the exits, and I guess the emergency shoots [sic] and so forth and so on.”).

The majority relies on the following passage from the district court’s order as evidence that the court somehow forgot appellants’ theory: “Apparently, the plaintiffs’ theory is that a fully loaded, fully fueled 747 aircraft, moving down a closed, under-construction runway at take-off speed in a typhoon, should be able to withstand a collision with construction cranes, bulldozers, and the like.” ER 85 (emphasis added). According to the majority, the claim so characterized relates to the “integrity of the aircraft,” whereas appellants are complaining about the adequacy of the “plane’s emergency and evacuation equipment.”

The distinction the majority draws is a false one. A plane that hits a stationary object will, of course, have its integrity impaired. The only question is whether the plane’s safety equipment will protect the passengers from sustaining injuries and enable them to escape to safety. The district court’s reference to whether the plane could “withstand” such a collision obviously pertains to whether the aircraft, though damaged, would be able to preserve the lives and safety of the passengers, and this would depend on whether the aircraft’s various features (e.g., emergency chutes, fire protection system, doors) were appropriately designed to sustain a collision yet still allow those inside to walk away unharmed. The district *997court’s statement, though perhaps inartful, thus reflected appellants’ theory that the aircraft’s safety features were not designed to withstand a collision.

Even if the district court mistakenly thought the suit against Boeing and Goodrich involved “crashworthiness,” rather than “the malfunctioning of the allegedly flawed design and manufacture of the plane’s emergency and evacuation equipment,” maj. mem. at 3,1 don’t see why this would matter. The majority claims that the differences in the two theories are material because the “evidence and witnesses” under these two theories would be different. Id. But the majority offers no clue as to why it believes that the “evidence and witnesses” would be different under one theory rather than the other. By whatever name, appellants’ theory of the case was that Boeing and Goodrich did not do enough to ensure the safety of the passengers in case of collision. The ability to evacuate quickly and safely is part-andpaxxel of the aircraft’s crashworthiness, and the “evidence and witnesses” under either theory would have been the same. Other than its conclusory assertion, the majority offers nothing to the contrary.

Next, the majority argues that “unlike the initial order denying Boeing’s original motion to dismiss, which names and rejects Singapore as the single alternative forum, the order upon reconsideration names three alternative forums: Singapore, Canada, and Taiwan.” Maj. mem. at 3. According to the majority, the district coxxrt “failed to balance the competing interests fairly by comparing the domestic forum to a particular foreign forum.” Id.

But the district court is not required to limit itself to a single foreign foram that it believes is best situated for appellants to bring their lawsuit. The district court clearly considered each of three fora preferable to the United States after weighing the relevant public and private factors. That the district court found three such fora, rather than one, only strengthens its analysis. It certainly does not weaken it.

Third, the majority argues that there is an inconsistency in the district court’s February 6 order between its unwillingness to try Singapore Airlines and appellees in a joint trial, “because to do so would risk prejudice to both defendants,” ER 86, and the district court’s conclusion that an alternative forum exists “in which the claims can be resolved in a single action against all defendants,” ER 91. This argument was never raised by appellants and, to the extent it was error, it was waived. See Walls v. Wells Fargo Bank, N.A., 276 F.3d 502, 511 (9th Cir. 2002).

In any event, the district coxxrt did not err because an “action” is not the same as a “trial.” The district court reasonably concluded at page 5 of its order that a joint trial would unduly prejudice appellees. At page 10, the district court was considering a different issue — not whether there could be a joint trial, but whether a single forum was available to resolve all the claims; it concluded that appellants’ claims could be resolved in a single action in a foreign forum. Because that single action could involve separate trials (thus avoiding the prejudice issue), there is no inconsistency in the district coxxrt’s order.

This is not only a plausible interpretation of the district court’s order, it is the only plausible interpretation; otherwise, a district court could never find another forum adequate once it concluded that the case was not capable of being resolved in a single trial. Had appellees been given an opportuxxity to respond to this argument— which they were not, since appellants never raised it — they surely would have pointed out the flaw in the majority’s homegrown rationale.

*998The majority’s remand is, in any event, pointless. The district court held that it would be more convenient to try this case in a foreign forum. In coming to this conclusion, the court weighed a variety of public and private factors, including (to name a few): (1) “the administrative difficulties flowing from court congestion”; (2) “the local interest in having the matter decided locally”; (3) “familiarity with governing law and avoidance of unnecessary problems in conflicts of law or application of foreign law”; (4) “the unfairness of burdening citizens in an unrelated forum with jury duty”; (5) that plaintiffs had already initiated suits against defendants in foreign fora; (6) that defendants had consented to jurisdiction in foreign courts; (7) “the relative ease of access to sources of proof’; (8) the “residenee[s] of the parties and witnesses”; (9) the “availability of compulsory process for attendance of witnesses”; (10) the “costs of bringing willing witnesses and parties to the place of trial”; (11) “access to physical evidence”; and (12) the “enforceability of judgments.” ER 89-92. The district court considered all these factors thoroughly, weighed the competing interests carefully and concluded that the case should be dismissed on forum non conveniens grounds. The three “errors” the majority purports to find cannot possibly affect this balance. On remand, the district court will take a deep bow to our ruling, correct the “errors” my colleagues have identified and reach exactly the same conclusion. We’ll be right back where we started — with a district court order of dismissal that we are bound to affirm under Piper Aircraft. The parties will have wasted at least two years for no good reason.