Students responded to a self-administered instrument that consisted of more than 750 questions. Starting in 1980, follow-up tracking and interviewing began to contact respondents, and follow-up data were collected primarily through personal interviews. Of the original 7,727 students for whom data were available, 5,145 or 66.6 percent completed follow-up interviews and questionnaires as adults. The study focused on early adolescent psychosocial development, problem behavior outcomes, and the resolution of nonmarital adolescent pregnancy. Findings confirmed previous research showing the impact of gender, race, and social class differences on problem behavior occurrence; empirical support for the causal model explanation; and limited evidence that psychosocial antecedents mediated the impact of demographic predictors of adolescent problem behavior. Specifically, social background characteristics were salient predictors of adolescent delinquent behavior, alcohol and drug involvement, and nonmarital pregnancy. Black adolescents were less engaged in delinquent behavior and drugs and more involved in nonmarital pregnancy than white adolescents. Effects of demographic predictors on problem behavior held for both early and late adolescence. Study limitations, areas for further research, and policy implications of the findings are noted. Appendixes contain ordinary least squares and logistic regression data on nonmarital adolescent pregnancy. 244 references, 33 tables, and 25 figures