Previous studies have shown that peer dysfunction in adolescence predicts depression in adulthood, even when controlling for certain individual- and/or family-level characteristics. However, these studies have not controlled for numerous potential familial confounders, precluding causal inferences. The present study therefore used a sibling comparison design (i.e., comparing siblings within families) to test whether peer dysfunction (e.g., lack of friendships, victimization) in adolescence continues to predict depression in adulthood after accounting for unmeasured familial confounds and individual characteristics in adolescence. Participants’ (N = 85) dysfunction with peers was assessed in adolescence (mean age = 13.21, SD = 3.47) by self- and parent-report, and adult depressive symptoms were assessed up to five times, up to 38 years later. Multilevel modeling was used to examine the effect of adolescent peer dysfunction on adult depressive symptoms after adjusting for familial confounds and/or individual characteristics in adolescence (e.g., baseline depressive symptoms, dysfunctional relations with siblings/parents). Both self-reported (b = 1.28, p < 0.001) and parent-reported (b = 0.56, p = 0.032) adolescent peer dysfunction were associated with greater depressive symptom severity in adulthood in unadjusted models. Self-reported (but not parent-reported) adolescent peer dysfunction continued to predict adult depressive symptoms after controlling for familial confounding and measured covariates such as adolescent depressive symptoms and relations with siblings and parents (b = 1.06, p = 0.035). Although confidence intervals were wide and the potentially confounding effects of numerous individual-level factors were not ruled out, these findings provide preliminary evidence that perceived peer dysfunction in adolescence may be an unconfounded risk factor for depressive symptoms in adulthood.