OBJECTIVE: The authors’ goal was to conduct an adult follow-up of subjects who had participated in a study of nortriptyline for childhood depression. METHOD: The study group represented 100 (90.9%) of the original 110 subjects and included 72 subjects who had a prepubertal diagnosis of major depressive disorder and 28 normal comparison subjects. Subjects were assessed with semistructured research interviews given by research nurses who were blind to the subjects’ original diagnoses. RESULTS: In the original study, the mean age of the children with prepubertal major depressive disorder was 10.3 years (SD=1.5); at adult follow-up the mean age of these subjects was 20.7 years (SD=2.0). At follow-up, significantly more of the subjects who had prepubertal diagnoses of major depressive disorder (N=24 [33.3%]) than normal comparison subjects (none) had bipolar I disorder. Subjects who had prepubertal diagnoses of major depressive disorder also had significantly higher rates of any bipolar disorder than normal subjects (48.6% [N=35] versus 7.1% [N=2]), major depressive disorder (36.1% [N=26] versus 14.3% [N=4]), substance use disorders (30.6% [N=22] versus 10.7% [N=3]), and suicidality (22.2% [N=16] versus 3.6% [N=1]). Parental and grandparental mania predicted bipolar I disorder outcomes. CONCLUSIONS: High rates of switching to mania have implications for the treatment of depressed children. The authors discuss the reasons for their finding a higher rate of bipolar disorder in this outcome study than was found in the one other adult outcome study of prepubertal major depressive disorder.