The current paper uses data from a longitudinal study of a high-risk sample to test the relation between adolescent alcohol and drug use and later young adult autonomy, positive activity involvement, and perceived competence. Participants (children of alcoholics and demographically matched controls) were assessed in three annual interviews in adolescence (mean age: 12.7 years at Time 1) and then again 5–7 years later, in young adulthood (median age: 20 years). Path analyses and latent growth curve models tested the effects of adolescent substance use on both self-reported and collateral-reported outcomes, controlling for correlated risk factors (parental alcoholism, adolescent psychopathology, and parental support), preexisting levels of the outcome, and concurrent young adult substance use. Results showed that adolescent drug use had a significant, unique negative effect on later autonomy and perceived competence. Alcohol use effects were more complex. Adolescent heavy drinking was associated with less positive adult outcomes, but more so in collateral reports than in self-reported outcomes. Moreover, young adult heavy drinking was either uncorrelated with or positively correlated with higher levels of perceived competence, suggesting different developmental significance of alcohol use in adolescence than in young adulthood.