The main objective was to analyse whether involvement in bullying at school predicts depression, and whether depression predicts involvement in bullying in middle adolescence. A total of 2,070 15-year-old girls and boys in two Finnish cities were surveyed at ninth grade (age 15) at schools, and followed up 2 years later in the Adolescent Mental Health Cohort Study (AMHC). Depression was measured by a Finnish modification of the 13-item short Beck Depression Inventory. Involvement in bullying was elicited by three questions focusing on being a bully, being a victim to bullying, and being left alone by peers against one's wishes. Similar questions were posed at both time points. Statistical analyses were carried out using cross-tabulations with chi-square/Fisher's Exact Test statistics, and logistic regression. The results summarized that, both being a victim to bullying and being a bully predicted later depression among boys. Among girls, depression at T1 predicted victimisation at T2. Depression at T1 predicted experience of being left alone at T2 among both sexes. It was concluded that victimisation to bullying may be a traumatising event that results in depression. However, depression also predicts experience of victimisation and of being left alone against one's wishes. Depression may impair an adolescent's social skills and self-esteem so that the adolescent becomes victimised by peers. However, depression may also distort and adolescent's experiences of social interactions.