OPINION OF THE COURT
Chief Judge Kaye.Plaintiff Ramsar Williams, an infant, was severely burned in a fire started by several other children with whom he was playing. Ramsar and his father brought this action against the children and their parents, as well as manufacturers and distributors of Ramsar’s clothing, which allegedly contributed to the burns.
One year before trial, plaintiffs settled with four defendants —Talon Corporation, Salant Corporation, Thompson Company and Travis Textiles — for a total of $900,000. During trial, plaintiffs settled with individual defendants Thomas and Robert Niske (the Niske defendants) for $100,000, and with defendant Union Underwear Company on a high-low basis, Union agreeing to pay a base amount of $100,000 and up to $400,000 more to assure that plaintiffs at trial recovered at least $500,000. Defendant Billy the Kid (BTK) did not settle. The jury returned a verdict of $2,600,000, apportioning liability 35% to BTK, 30% to Niske and 35% to Union. No proof was offered with respect to the percentage fault (i.e., the equitable share of the damages) of the four defendants who had settled before trial.
Under General Obligations Law § 15-108 (a), the settlements unquestionably reduce BTK’s liability for the verdict. The question before us is how much that reduction should be.
Supreme Court concluded that BTK’s liability should be reduced to $10,000, leaving Union then responsible for $490,000. The Appellate Division modified, concluding that BTK must pay $595,000, and Union thus only $100,000. BTK appeals and plaintiffs cross-appeal by leave of the Appellate Division.
*440I.
Analysis begins with General Obligations Law § 15-108 (a):
"When a release * * * is given to one of two or more persons * * * claimed to be liable in tort for the same injury * * * it does not discharge any of the other tortfeasors from liability * * * but it reduces the claim of the [plaintiff] against the other tortfeasors to the extent of any amount stipulated by the release or the covenant, or in the amount of the consideration paid for it, or in the amount of the released tortfeasor’s equitable share of the damages under article fourteen of the civil practice law and rules, whichever is the greatest.”
Where there is only one settling defendant whose equitable share of the damages has been determined — the situation described in the statute — calculating the amount to be paid by nonsettling defendants is a simple task. If the settlement figure is below the settling defendant’s equitable share, the nonsettling defendants pay only their equitable share of the verdict. If the settlement exceeds the settling defendant’s equitable share, the nonsettling defendants pay only the difference between the settlement and the verdict, equitably apportioned among them. If the settlement exceeds the verdict, the nonsettling defendants have no liability at all.1
Here, however, the situation is more complicated. There are six settling defendants, four of whom settled before trial with no assessment of their equitable share of the damages. In these circumstances, the statute cannot be applied literally. First, by its terms the statute contemplates a release of only one defendant; here, releases were given to several. Second, the statutory formula requires a comparison of equitable share with actual settlement payment. Where four defendants settled without any assessment of their equitable share, it is impossible to make the comparison as to them.
*441The parties suggest three methods for calculating the amount by which BTK’s liability is to be reduced.
The first method, advocated by BTK and applied by Supreme Court, begins with the defendants who settled during trial; their equitable shares were determined, and the statutory comparison thus can be made as to them. According to this method, first the $2,600,000 verdict is reduced by the 65% equitable liability ($1,690,000) of the defendants who settled at trial, that being greater than the amounts they actually paid.2 The resulting amount — $910,000—is then reduced by the $900,000 plaintiffs obtained in the pretrial settlements. BTK is thus liable for $10,000, and Union for $500,000; plaintiff would recover a total of $1,510,000.
The second method, urged by Union and applied by the Appellate Division, begins by subtracting the $900,000 pretrial payments from the verdict ($2,600,000 — $900,000 = $1,700,000), thus treating the case as if total liability were $1,700,000 instead of $2,600,000 and doing the statutory comparison as to the remaining defendants. If the remaining defendants had all proceeded to a verdict in the amount of $1,700,000, their shares would have been $595,000 from BTK (35%), $595,000 from Union (35%) and $510,000 from Niske (30%). Those shares being greater than the actual settlement amounts paid by Union and Niske, BTK’s liability is its equitable share of the reduced verdict — $595,000—and Union’s $100,000. Plaintiffs recover a total of $1,695,000.
The third method, urged by plaintiffs, would total the equitable shares of the settling defendants (65% or $1,690,000, allocating nothing for the defendants who settled before trial, since we do not know what their equitable share should be), then total the settlement payments ($600,000 + $900,000 = $1,500,000), and reduce the verdict ($2,600,000) by the greater aggregate number ($1,690,000). This results in a liability for BTK in the amount of $910,000; Union is responsible for $100,000; and plaintiffs recover a total of $2,010,000.
Of the three methods, we conclude that the Appellate Division’s calculation most closely approximates the statutory language and purposes, and we therefore affirm.
*442II.
General Obligations Law § 15-108 was enacted in 1972 to encourage settlements by abrogating the common-law rule that a release of one joint tortfeasor automatically released all others — a rule perceived as trapping litigants who would naturally assume that settlement with one party would have no effect on their ability to pursue another (see, Hill v St. Clare’s Hosp., 67 NY2d 72, 83; 1972 NY Legis Doc No. 65[K], reprinted in 1972 McKinney’s Session Laws of NY, at 3237-3239). Abrogation of the rule encouraged plaintiffs to settle by ensuring preservation of their cause of action against others. To address the specter of double recovery by plaintiffs, the statute as originally enacted codified the rule that a release would reduce the plaintiff’s claim against the other tortfeasors "to the extent of any amount stipulated by the release or the covenant, or in the amount of the consideration paid for it, whichever is the greater” (see, L 1972, ch 830; 1972 McKinney’s Session Laws of NY, at 3239).
At common law, there was no right to contribution from wrongdoers not part of the lawsuit, including defendants who settled (see, Fox v Western N. Y. Motor Lines, 257 NY 305, 308). General Obligations Law § 15-108 did not change this rule, and thus the statute did not need to include any inducement for defendants to settle — release from further liability and freedom from contribution was inducement enough. (12th Ann Report of Jud Conf on CPLR, reprinted in 1974 McKinney’s Session Laws of NY, at 1816-1817.)
General Obligations Law § 15-108, however, failed to take into account our decision in Dole v Dow Chem. Co. (30 NY2d 143) — decided just as the statute was enacted — which significantly changed the law of contribution by requiring apportionment of liability among all tortfeasors. While encouraging settlements remained an important goal of the judicial system, Dole-Dow announced another important goal: equitable loss-sharing among wrongdoers (Dole v Dow, 30 NY2d at 153). To further this goal, Dole-Dow was interpreted to allow defendants who proceeded to judgment to obtain contribution from all tortfeasors, even those who had settled (see, Codling v Paglia, 32 NY2d 330, 344).
The Judicial Conference thereafter commissioned a study on the effects of Dole-Dow, including its effects on settlement. The study concluded that Dole-Dow provided a disincentive for defendants to settle since they could no longer be assured that *443a settlement would include immunity from liability, and recommended amending General Obligations Law § 15-108 to foreclose actions for contribution against settling defendants. That objective was accomplished with the addition of General Obligations Law § 15-108 (b).
Consistent with the goals of Dole-Dow, however, the study concluded that the statute also should ensure that nonsettling defendants did not bear more than their equitable share of a plaintiff’s damage (see, Occhialino, Contribution, reprinted in 19th Ann Report of NY Jud Conf, at 217, 224). That objective was accomplished by amendment to subdivision (a) (see, Report on Right of Contribution, 1974 NY Legis Ann, at 15-17). Under the amended subdivision (a), even if the settlement amount is less than the settling tortfeasor’s equitable share, the nonsettling defendants pay no more than their equitable share of the verdict. If the settlement amount exceeds the settling defendant’s equitable share, the benefit inures to the nonsettling defendants, who are credited with the larger amount actually paid (1974 NY Legis Ann, at 17; see also, Killeen v Reinhardt, 71 AD2d 851; Werner v Our Lady of Lourdes, 60 AD2d 791).
This rule has been criticized as a disincentive to settlement because it can result in undercompensation to a settling plaintiff and a windfall to a nonsettling defendant (see, Carbone, Repealing New York’s Post-Settlement Equitable Share Reduction Scheme: An Idea Whose Time Has Come, 49 Alb L Rev 856, 864-867; Green, General Obligations Law Section 15-108: An Unsettling Law, 55 [No. 6] NY St BJ 28, 29-30; Occhialino, op. cit., at 224). It is clear, however, that in amending subdivision (a), the Legislature was well aware of such concerns (see, Occhialino, op. cit., at 219), but saw the risk of chilling plaintiff’s desire to settle as largely outweighed by the benefits of early settlement, and the potential windfall to defendants as necessary to ensure that plaintiffs do not in effect have a double recovery against nonsettling defendants (1974 McKinney’s Session Laws of NY, at 1818 [citing Kelly v Long Is. Light. Co, 31 NY2d 25, 30]; see also, Purcell v Doherty, 102 Misc 2d 1049, affd 80 AD2d 755, affd 55 NY2d 985).
Thus, while the general purpose of General Obligations Law § 15-108 (a) is to encourage settlements, the statute is concerned as well with assuring that a nonsettling defendant does not pay more than its equitable share. This consideration is *444particularly pertinent in evaluating the three proffered methods, to which we next turn.
III.
As the Appellate Division observed, given the numerous possible variations in multidefendant litigation, the task of formulating a single method for reducing verdicts to account for settlements "appears all but insuperable if the statute’s important objectives are not to be sacrificed in the process * * * In the end then, it will be up to the courts to determine the method of verdict reduction which best promotes the statute’s broad objectives.” (181 AD2d 307, 312.) Here, we must determine the most appropriate method where equitable liability has been determined for some settling defendants but not for others. While none of the three proffered methods is a perfect solution, we conclude that the method advocated by Union and adopted by the Appellate Division most closely comports with the statutory scheme.
We first reject the method suggested by plaintiffs, which meets neither the statutory language nor its purposes. By its terms, the statute directs that the reduction be the greater of actual settlement payment or equitable share. While plaintiffs purport to make that precise analysis — aggregating all six settling defendants and their payments — it is a false comparison because in fact no equitable share was determined as to the four defendants who settled before trial. As a practical matter, moreover, the result is that — contrary to the statutory objective — BTK is deprived of credit for the $900,000 pretrial settlement.
Unlike plaintiffs’ method, the remaining methods both make the statutory comparison of equitable share against actual settlement payments for those settling defendants. Both, in other words, compare like figures, whereas plaintiffs do not. Where the remaining two methods differ, however, is in their treatment of the $900,000 pretrial payment: BTK and Supreme Court would first deduct the 65% and then the $900,000; Union and the Appellate Division would first deduct the $900,000 and then the 65%.
We reject BTK’s proposed method. The jury apportioned fault (65%-35%) only among the trial defendants, not including the pretrial settling defendants; the pretrial settling defendants were not before the jury. BTK’s method, however, changes the jury’s allocation by deducting 65% of the total *445verdict, which includes the $900,000 paid by the pretrial settling defendants. BTK is not entitled to both a percentage reduction of the $900,000 and subtraction of the $900,000.
The Appellate Division’s method ameliorates that problem. Plaintiffs received $900,000 up front, which reduced the verdict. Plaintiffs’ remaining uncompensated damages — $1,700,000—were then apportioned according to the equitable fault assigned by the jury, with each nonsettling defendant responsible only for its equitable share. That method best accomplishes the purposes of General Obligations Law § 15-108 (a).
Accordingly, the order of the Appellate Division should be affirmed, without costs.
Judges Simons, Titone, Hancock, Jr., and Bellacosa concur; Judge Smith taking no part.
Order affirmed, without costs.
. Even though a defendant in a multidefendant suit settles, proof as to the settler’s fault may still be presented at trial and the settler’s equitable share determined. Generally, the remaining defendants will choose to present such proof as a way of limiting their own equitable share (see, Siegel, NY Prac § 176, at 267 [2d ed]; McLaughlin, Practice Commentaries, McKinney’s Cons Laws of NY, Book 23A, General Obligations Law § 15-108, at 701). The parties do not raise, and we therefore do not consider, where the burden for pleading and proving the equitable fault for settling defendants lies.
. For convenience, throughout this opinion we consider the 30% and 35% allocated to the defendants who settled during trial as a single 65% figure. Since each of those defendants settled for less than their equitable share, their percentages can be aggregated without changing the result. We make no determination as to whether it is proper to aggregate amounts or make the calculations on a defendant-by-defendant basis.