Understanding the interaction of peer problems with substance use influence on psychiatric symptoms has important clinical implications. We tested the hypothesis that peer problems moderate the relationship between substance use and psychiatric symptoms on a sample of 280 adolescents receiving psychiatric treatment. Age, race, and gender served as covariates. Our model accounted for a significant proportion of the variance in psychiatric symptoms (R2 = .379, p < .001). A significant moderating effect of peer problems x substance use on psychiatric symptoms was found (p < .05). Next, we estimated the conditional effect of substance use on psychiatric symptoms as a function of low, moderate, and high levels of peer problems. Significant effects for low-level peer problems emerged, such that as peer problems increase by one unit, the effect of substance use on psychiatric symptoms increases by one half unit. Implications for peer-based interventions are discussed.