The body of research on the development of antisocial behavior includes a large number of longitudinal panel studies covering various segments of the life course and a diversity of variables, and a larger number of supporting cross-sectional studies. While there is much still to be learned from additional research on this topic, there is also value in synthesizing and integrating what is already known. The complexity and diversity of this research, however, makes that a formidable task, one which has not yet been cogently accomplished. This project proposes to use advanced meta-analysis techniques to achieve a quantitative integration of the empirical research findings that bear on three issues: (1) The nature and extent of the continuities and discontinuities from early childhood to adulthood in the development of antisocial behavior; (2) the relative independent strength of association of risk and protective variables with persistence and desistance in antisocial behavior; and (3) the generalizability of the relationships on these issues across populations and cohorts.