This study used longitudinal survey and social network data covering sixth through ninth grades to test whether internalizing symptoms make early adolescents more prone to (1) exposure to and (2) influence by substance‐using peers. Random effects regressions revealed that increases in symptoms were significantly associated with increases in the proportion of friends who used cigarettes, alcohol, and marijuana; some associations weakened across grades. Event history models revealed that the effect of friends' smoking on smoking initiation decreased as internalizing symptoms increased; symptoms did not moderate the effects of friends' alcohol and marijuana use on alcohol and marijuana use initiation. These findings counter the influence hypothesis of the co‐occurrence of internalizing symptoms with substance use and partly support the exposure hypothesis.