This Article's normative claim - that a rule allowing subsequent challenges to class action settlements is compelled by our basic intuitions of fairness and justice when class members could not conceivably have agreed to the arrangement had they been present but not known their precise in the class - builds on the Rawlsian construct of fairness as the product of (hypothetical) decision-making in an behind a of ignorance, and the economics of human decision-making under conditions of uncertainty. This approach suggests that certain types of settlements in both high-individual-stakes/toxic torts/personal injury class actions and small-individual-stakes/consumer fraud class actions should be subject to subsequent challenge. First, this combined Rawlsian/economics analysis strongly suggests that all class settlements that provide for the possibility that any class members will receive negative relief, as in the Bank Boston litigation, are unfair and should be subject to challenge on adequacy of representation grounds. Second, with regard to high-individual-stakes class actions, the approach also suggests that settlements that create the possibility that some class members will receive no relief always should be subject to subsequent challenges. Third, again in high-individual-stakes class actions, subsequent challenges should be permitted with respect to settlements that provided all class members some relief, but that grossly deviated from a principle of equal payment for equal harms without investing administrative cost savings in the improvement of the of the most severely-injured such that the most-severely injured receive more than they would have under an equal- compensation-for-equal-injuries formula. Fourth, even in small individual stakes litigation, subsequent challenges should be permitted to settlements that provided for the possibility of providing zero compensation to any class members or that deviated from an equal-compensation-for-equal-injuries approach compensating without thereby reaping significant administrative cost savings that are dedicated to increasing the overall compensation pool for class members. The essence of the Rawlsian approach is a thought experiment regarding the ordering of society as a whole. In the thought experiment, Rawls postulates the presence of human beings under certain conditions - position conditions - and then reflects on what arrangements or rules those individuals would agree to as fair for the distribution of goods and entitlements in the society as a whole. The conditions Rawls sets for his thought experiment - individual decision-making, ignorance on the part of each individual as to their morally irrelevant or contingent characteristics beyond the veil, very high stakes for individual welfare and life prospects for the decision-makers, a generally shared moral sense of the fundamental equality of human beings - readily translate from original (persons deciding on the rules for social ordering as a whole) to the toxic tort/products liability class action original position, in which class members must choose a distribution regime for compensation for possible current and future cases of disabling or even fatal diseases, conditions, or injuries. Thus, if persons in Rawls' original adhere to a maximin principle of avoiding worst possible outcomes, and that adherence deserves normative weight, we should expect that class members in toxic tort/products liability cases behind a veil of ignorance will adhere to a maximin approach, and we should accord normative weight to that adherence. The Article does not rely solely on the extension of the Rawlsian original thought experiment to the class action context, however. One of the predictions of neoclassical economic theory, as well as a basic finding of behavioral/experimental/empirical economics is risk aversion in human decision-making in the absence of an ability to self- or third-party-insure against bad outcomes. Other findings are an aversion to unequal distributions of wealth and other goods absent some objective justification for inequality, and an aversion to prospective losses (as opposed to gains or foregone gains). These well-established findings, in and of themselves, suggest that settlements of the sort at issue in Stephenson and Homeside could not have garnered the agreement of class members had they been able to give or deny their consent. In addition, surveys I conducted of first-year law students demonstrate a strong hostility to settlements that entail the risk that a class member might be left without any relief for severe injuries, and a strong predisposition toward settlements that ensure equal outcomes for equally harmed class members.