The proposed set of explanatory variables includes those drawn from the biological, psychological, and social sciences. The measures of delinquency and crime, the primary dependent variables, include both self-reported criminality and officially recorded criminal acts. The design is prospective, with 11 annual data waves, supplemented by a series of smaller integrated studies designed to pursue special issues at greater depths over shorter time intervals. The latter studies will use a variety of data collection and measurement approaches (e.g., observation, telephone interviews, questionnaires, ethnographic descriptions). The sample for both the longitudinal and special studies will be a national probability sample, involving multiple birth cohorts, stratified for risk of serious involvement. The sampling plan will provide projected sequential data covering a 30-year span (age 5 to 35). The proposed research design will permit a genuine age-cohort analysis and the possibility of isolating age, cohort, and period effects. 8 footnotes and approximately 300 references.