The current study tested parent alcoholism effects on growth curves of adolescent substance use and examined whether parent and peer influences, temperamental emotionality and sociability, and stress and negative affect could explain parent alcoholism effects. Longitudinal latent growth curve modeling showed that adolescents with alcoholic fathers, boys, and adolescents with drug-using peers had steeper growth in substance use over time than did adolescents without alcoholic fathers, girls, and adolescents without drug-using peers. Data were consistent with father's monitoring and stress as possible mediators of paternal alcoholism effects. However, the direct effects of paternal alcoholism on substance use growth remained significant even after including the hypothesized mediators in the model. This suggests that other (unmeasured) mediators are necessary to fully explain paternal alcoholism risk for adolescents' escalating substance use over time.