Opinion by Judge FLETCHER; Dissent by Judge BOOCHEVER.

Plaintiffs Fuku-Bonsai, Inc. and its president, David Fukumoto, (collectively, “Fuku-Bonsai”) appeal from the district court’s order dismissing its complaint because it was barred by release language in a settlement agreement between the parties. We have jurisdiction pursuant to 28 U.S.C. § 1291,1 and we reverse.

I. BACKGROUND

Fuku-Bonsai is a Hawaii corporation that raises ornamental plants for commercial sale. Like many other commercial growers and nurseries, Fuku-Bonsai used a DuPont fungicide called Benlate. Growers suspected that the Benlate was contaminated with the herbicide atrazine after the Benlate destroyed their plants.2 In 1992, Fuku-Bonsai brought a products liability action against DuPont. It was one of over seventy Benlate cases filed in the Third Circuit Court of Hawaii that were consolidated for discovery. In 1994, before trial, Fuku-Bonsai settled its claims against DuPont for $2.5 million. As part of the settlement agreement, Fuku-Bonsai released DuPont from certain claims. The scope of that release is at issue in this appeal.

Shortly after Fuku-Bonsai settled and dismissed the product liability action, courts hearing Benlate cases in several *1033jurisdictions uncovered evidence that DuPont had systematically concealed “smoking gun” evidence in Benlate cases around the country. Other Hawaii plaintiffs, whose case had been consolidated with Fuku-Bonsai for discovery purposes, went to trial and won a verdict of $23 million for crop damages resulting from defective Benlate. See Kawamata Farms, 86 Hawaii at 222, 948 P.2d at 1063. That case helped expose DuPont’s widespread concealment of evidence in the Benlate cases, and revealed that DuPont had systematically withheld evidence proving that Ben-late was contaminated. DuPont was fined $1.5 million for fraudulently withholding evidence. Id. at 231, 948 P.2d 1055. The Hawaii Supreme Court deemed DuPont’s pattern of discovery abuse “an unusual, unique example of unprecedented discovery fraud [perpetrated] against the court.” See id. at 257, 948 P.2d 1055.

• After learning of DuPont’s fraudulent behavior, Fuku-Bonsai, like many other settling plaintiffs who had been misled by DuPont, in 1997 filed an action seeking damages and sanctions against DuPont and other defendants. Fuku-Bonsai claimed that DuPont fraudulently withheld or concealed information and induced it to settle its claims for less than fair value. We must determine whether the release Fuku-Bonsai signed as part of the settlement agreement forecloses its present claim that the settlement was induced by fraud.

DuPont contends, and the district court agreed, that the release bars the present action. DuPont insists that plaintiffs should have rescinded the settlement agreement and that having failed to tender the settlement proceeds, plaintiffs forfeited their right to sue. The district court, relying in part on an unpublished district court Benlate ruling, Matsuura v. Alston & Bird and E.I. DuPont de Nemours and Co., Inc., D.C. CV-96-01180-DAE (D.Haw.), involving identically worded releases signed by other Hawaii growers, concluded that under Delaware law,3 the release language barred plaintiffs’ claims and that a plaintiff who has settled must affirm or rescind a release contract in its entirety.

In an opinion filed after plaintiffs appealed, this court reversed the district court decision in Matsuura, holding that plaintiffs’ fraudulent litigation claims fell outside the scope of the release language. See Matsuura v. Alston & Bird and E.I. duPont de Nemours and Co., Inc., 166 F.3d 1006 (9th Cir. 1999)(per curiam). Matsuura is indistinguishable from this case. We reverse the district court and hold that the release signed by Fuku-Bonsai does not bar its fraudulent inducement claim against DuPont.

II. ANALYSIS

In Matsuura, we held that, under Delaware law, plaintiffs who have been fraudulently induced to settle tort claims have a choice of remedies: they may rescind the contract or they may affirm the contract and sue for fraud. See Matsuura, 166 F.3d at 1008; see also DiSabatino v. United States Fidelity & Guar. Co., 635 F.Supp. 350 (D.Del. 1986). The DiSabatino court had noted that “the Delaware courts are in accord with the basic contract principle that a party defrauded on a contract may elect either to rescind the contract or to affirm it and sue for damages.” DiSabatino, 635 F.Supp. at 352. It predicted that “a Delaware court would rule that a tort claimant has an election to stand on a fraudulently induced release and proceed on a cause of action based on fraud.” Id. In Matsuura, we concluded that “the policy and legal analysis” of DiSabatino applies “regardless of who corn-*1034mitted the fraud.” Matsuura, 166 F.3d at 1008. We rejected DuPont’s argument that a defrauded tort plaintiff was limited to the remedy of recission. See id. at n. 4.

In Matsuura, we looked at the scope of the release signed by Matsuura. We reasoned that “the Supreme Court of Delaware would not interpret the Matsuura-DuPont releases to bar a claim of fraudulent inducement of the releases themselves.” Matsuura, 166 F.3d at 1009. We concluded that Delaware principles of contract construction did not support DuPont’s broad reading of the release. Under Delaware law, when specific recitals in a release are followed by general language, the specific language restricts the scope of the general release language. See Adams v. Jankouskas, 452 A.2d 148, 156 (Del. 1982). If an apparent conflict exists between the specific release language and the terms of the general release, Delaware relies on a rule of construction that “words of general application used in the release which generally follow a specific recital of the subject matter concerned are not to be given their broadest significance but will be restricted to the particular matters referred to in the recital.” See id.

The Fuku-Bonsai release is identical to the one in Matsuura. No facts that would distinguish our case from Matsuura have been suggested or revealed in the record before us. Its reasoning squarely applies to this case.

Although the Delaware Supreme Court has not yet ruled on the effect of the Benlate releases,4 we note that the Delaware Court of Chancery recently recognized the “conclusive nature” of the evidence of fraudulent misrepresentation and concealment in Matsuura when distinguishing insubstantial fraud allegations in the case before it. See In re U.S. Robotics Corp. Shareholders Litigation, 1999 WL 160154 at *12, n. 3 (Del.Ch., Mar.16, 1999). That Delaware court noted that in Matsuura,

[t]he plaintiffs supported their fraudulent inducement claim with evidence which had previously convinced two federal district courts that the defendants had misrepresented the nature of certain material evidence and had concealed other evidence plaintiffs sought. The conclusive nature of the evidence in Matsuura is far different than the sketchy, implausible, and insubstantial fraud allegations that movants advance in this case.

III. CONCLUSION

We hold, as we did in Matsuura, that plaintiffs’ claim for fraudulent inducement is not barred by the release language in the product liability settlement agreement. DuPont’s fraudulent actions and lack of good faith negotiations undermined plaintiffs’ ability to bargain freely for a fair settlement. We reiterate the observation made in Matsuura that Delaware’s policy of favoring the voluntary settlement of legal disputes is advanced by allowing Fuku-Bonsai a remedy for its victimization in settlement negotiations.

REVERSED AND REMANDED.

dissenting:

It is agreed that Delaware law applies in deciding whether the release signed by Fuku-Bonsai bars it from bringing this suit claiming that DuPont fraudulently withheld or concealed information that induced Fuku-Bonsai to settle its claims for less than fair value.

When we decided Matsuura v. Alston & Bird and E.I. du Pont de Nemours and Co., 166 F.3d 1006 (9th Cir. 1999) (per cu-riam), the Delaware Supreme Court had not accepted certification of the controlling state law question. Once the Delaware Supreme Court answers that question, our decision in Matsuura will no longer be binding on the interpretation of Delaware *1035law. The Delaware Supreme Court, not a federal court, will resolve that state law issue.

Because, as indicated in the majority opinion, the Delaware Supreme Court has accepted certification of this question in a similar case, see E.I. du Pont de Nemours and Company v. Florida Evergreen Foliage and Louis Chang, 98-2242-CIV-GOLD, I would await its decision before attempting to decide the Delaware law question. The added delay should not prove too great a hardship as Fuku-Bon-sai has retained the $2.5 million it received in settlement of its claims.