In a study of a total high-school population, 2300 students aged 16-17 years were screened for depression (BDI, CES-DC). Those with a self-evaluation indicating depression, together with controls matched for sex, age, and class were interviewed (DICA-R-A). The 177 pairs, where both individuals were interviewed and the control had no lifetime diagnosis of depression, were analysed in the form of paired differences for psychosocial factors and compared within diagnostic groups. The psychosocial factors were measured with the ISSI subscales and six attitude questions about family climate (KSP). Adolescents with an episode of major depression during part of the last year did not differ from their controls. Those with long-lasting depressive symptoms, i.e. dysthymia with or without episodes of major depression, had a more limited social interaction and were not satisfied with it. They also evaluated their family climate and attachment network as being more inadequate than did their controls. Depressed adolescents with comorbid conduct disorder had a more negative evaluation of availability and adequacy of both social interaction and attachment network than their controls. This group had a very negative view of their family climate. Since this is a case-control study conclusions cannot be drawn about cause and effect.