OBJECTIVE This study examines the main and interaction effects of known social risk factors for substance use (inadequate parental monitoring and substance using friends) in adolescence and a polygenic score in predicting marijuana and tobacco use in adolescence and young adulthood. METHODS Phenotypic and genetic data were derived from a longitudinal study of a cohort of urban, predominately African American youth. Last year substance-use measures were collected annually from 8th grade through age 22. Participant self-reports of substance-using friends and parent monitoring were obtained yearly from Grades 8 to 12. Using longitudinal latent class analysis, the authors identified parallel developmental trajectories of tobacco and marijuana use and parent monitoring and the proportion of substance-using friends. RESULTS Two trajectories were identified for tobacco and marijuana use, characterized by moderate versus little-to-no use. Additionally, 2 latent profiles were found for the social environment profiles: those characterized by higher parent monitoring and a lower proportion of substance-using friends versus lower parent monitoring and a higher proportion of substance-using friends. CONCLUSIONS We found main and interaction effects for the polygenic score and social environment profile in predicting the longitudinal classes of marijuana and tobacco use. With respect to the interaction effect, membership in the moderate-use classes of marijuana and tobacco use was highest among those in the social environment profile characterized by lower parent monitoring and a higher proportion of substance-using friends.