Background: Adults with depressive disorder report high rates of sub-optimal maternal care in childhood. Despite the greater salience of relationships with parents earlier in life, associations with parenting style have not yet been systematically studied in adolescent onset disorder. Methods: A six-wave, 3-year study of adolescent health in 2032 Australian secondary school students provided an opportunity to undertake a two-phase study of early onset depression. Between waves 2 to 6, a self-administered computerised form of the revised Clinical Interview Schedule (CIS-R) was used to generate a first phase diagnosis of ICD-10 depressive episode. Each subject with a CIS-R-defined depressive episode was selected for second phase assessment together with two subjects from the CIS-R non-cases in each school. Second phase assessment included a second diagnostic assessment using the depression and hypomania modules of the Composite International Diagnostic Interview (CIDI) and assessment of paternal and maternal style using the Parental Bonding Instrument. Results: A total of 1947 out of 2032 subjects in the sampling frame (95.8%) participated in the cohort study (phase 1) at least once; 406 (94%) of the 435 selected subjects completed second phase assessment. One hundred and nineteen subjects fulfilled criteria for depressive episode on the CIS-R at one or more waves. Over the 30-month study period, 69 subjects (10 male, 59 female) fulfilled both CIS-R and CIDI definitions of depression at the same wave and were classified as 'definite depressive disorder'. Low maternal and paternal care held independent associations with both definitions of depression, with the effects clearest in those in the lowest quartile of reported care. After adjusting for low parental care, the associations between high parental control and depression were small. Conclusions: Sub-optimal parenting is associated with depressive disorder in adolescents. Low maternal and paternal care are each associated with a two- to three-fold higher rate of depressive disorder. These findings are consistent with an effect of sub-optimal parenting on the onset rather than course of disorder. Whether sub-optimal parenting is associated with a risk for the onset of depression outside the adolescent years has yet to be clarified.