John R. Snarey Laboratory of Human Development Harvard University Over the past 15 years, children and adults around the world have been asked if Heinz should steal a drug to save his dying wife, if Njoroge should disobey the rules to help a lost child, or some other similar moral dilemma. These cross- cultural studies have been undertaken to test Lawrence Kohlberg's theory, which posits a universal model of moral development. This review identifies the major empirical assumptions underlying Kohlberg's claim for cross-cultural universality, including culturally diverse samplings, universal moral questions, invariant stage sequence, full range of stages, and general applicability of the stages. It then reviews the cross-cultural research literature, much of which has not been previously published, and evaluates the support for each assumption. In addition to providing striking support for the underlying assumptions, the 45 studies examined here also identify some major caveats regarding the range and general applicability of the stages across cultures. In particular, biases in favor of complex urban societies and middle-class populations are identified. Based on these findings, the conclusion presents an alternative to Kohlberg's perspective on the relation between culture and moral development. Lawrence Kohlberg's stage model of moral development, briefly summarized in Table 1, has attracted a great deal of positive attention from psychologists and educators. Not sur- prisingly, however, his work has also inspired considerable criticism and revisionism (cf. Gibbs, 1977; Gilligan, 1982; Kurtines & Grief, 1974; Rest, 1983). The aspect of Kohl- berg's theory that has been most difficult for many social scientists to accept is the claim that the of moral reasoning about the social environment follows a uni- versal invariant sequence, toward the same universal ethical principles, in all cultural settings (cf. Bloom, 1977; Buck-Morss, 1975; Edwards, 1975, 1982; Guidon, 1978; Shweder, Preparation of this review was supported in part by National Institute of Mental Health Grant MH14088. I gratefully thank Agusto Blasi, John Broughton, John Gibbs, Lawrence Kohlberg, Stuart Hauser, Betty J. House, Harry Lasker, Robert LeVine, Joseph Reimer, Richard Shweder, and Carol Snarey for their helpful comments on a preliminary draft of this article. Special thanks to Carolyn Edwards for her extensive and careful comments. Requests for reprints should be sent to John Snarey, Laboratory of Human Development, Harvard University, Larsen Hall 300, Appian Way, Cambridge, Massachusetts 02138. 1982a, 1982b; Simpson, 1974; Sullivan, 1977). This article identifies the primary em- pirical assumptions underlying Kohlberg's claim of cross-cultural universality and clar- ifies the appropriate evidence necessary to judge the claim. It then presents a compre- hensive examination of the available empirical evidence that has accumulated over the last 15 years and evaluates the support or lack of support indicated. The assumptions are dis- cussed in evaluative order, from those that receive the most support to those that receive the least support. Assumptions and Hypotheses Kohlberg (1971) stated his claim for the cross-cultural universality of moral develop- ment as follows: All individuals in all cul- tures use the same thirty basic moral cate- gories, concepts, or principles, and all indi- viduals in all cultures go through the same order or sequence of gross stage development, though they vary in rate and terminal point of development (1971, p. 175). I believe that Kohlberg's claim implies at least five empir- ical assumptions, each of which may be un- derstood as a testable hypothesis. The first assumption is that moral devel- 202