Previous research into the psychopathology of children has indicated that genetic, biological, and psychosocial factors converge in the development of psychopathology. However, the majority of these studies have relied on cross-sectional data. The current study involved a longitudinal analysis of the association of baseline variables with psychiatric disorders in adolescents at baseline and at a 6-year follow-up. It was expected that the presence of psychological stress during the course of adolescent development would increase the risk of vulnerability to psychiatric disorders. Participants were 3,419 seventh through ninth graders who completed the Center for Epidemiological Studies Depression Scale; top scores and a random sample of the remainder of the students (n=581) were interviewed using the Schedule for Affective Disorders and Schizophrenia for School-Age Children. Follow-up interviews were completed with 490 of the participants 6 years later. Results indicated that the presence of psychiatric disorders decreased from 19.9 percent at baseline to 5.7 percent at follow-up. Multivariable regression analyses further indicated that baseline stressful life events and low family cohesion were associated with the presence of any disorder and affective disorder at baseline; not living with both parents at baseline increased the chances for affective disorder at baseline and follow-up. These findings remained significant after controlling for race, gender, and socioeconomic status. Finally, at both baseline and follow-up, anxiety and affective disorders were associated with being White and anxiety disorder was associated with being female. Given that family structure is a marker for an array of disorders, clinical interventions designed to reduce family stress are indicated. Tables, references