Depression characteristics were investigated in 507 adolescents using a two-stage longitudinal study design. The three-year longitudinal study started when the girls and boys were 11 and 12 years, respectively. In the initial (screening) phase the Children's Depression Inventory (CDI) was used to select subjects for phase-II of the study in which the Children's Depression Rating Scale-Revised (CDRS-R) interview was used. The caseness definition of Major Depression (MD) and Dysthymia (D) was based on the DSM-III-R criteria. Pubertal development was assessed by Tanner's staging. Estimated prevalence of Major Depression in the female sample for each increasing year of age was 2.2%, 2.7% and 4.1%. In the male sample the prevalence for the three years was 0.9%, 0.3% and 0.6%. There was no relation to age. The estimated mean prevalence of Dysthymia was 1.4% in girls and 0.8% in boys. No association between pubertal development and depression was found. Longitudinal data support the notion of chronic depression in early adolescence.