The high-risk youth (n=57), the youths with major depressive disorder (n=71), and the healthy control youths (n=48) and their families were recruited from 1987 to 1996. Subjects' lifetime and present DSM-III-R psychiatric symptomatology was assessed with the Schedule for Affective Disorders and Schizophrenia for School-Age Children-Epidemiologic Version and the K-SADS (6-18 years) Present Episode Version, with both the youth and parents or guardians being informants. The Psychosocial Schedule was administered to mothers by experienced interviewers. This is a semistructured interview that assesses the mother-child, father-child, peer, and marital relationships and the child's school performance. Except for 16 children who had disruptive disorder, the high-risk children were free of psychopathology. Overall, high-risk and healthy controls had similar psychosocial functioning. Marital relationships were worse in high-risk children with psychopathology. Youths with major depressive disorder had significantly more psychosocial problems and school difficulties for most domains measured compared with high-risk youth and healthy controls. Similar results were found when controlling for age, pubertal stage, race, sex, family composition, current and lifetime parental depression, and current and lifetime parental nonmood psychopathology. The authors advise that although family dysfunctional patterns apparently are related mainly to the child's depressive symptoms, longitudinal studies are required to establish causality. 2 tables and 50 references