Objective: Prior studies on antidepressant use in late adolescence and young adulthood have been cross-sectional, and prospective associations with childhood psychiatric problems have not been examined. The objective was to study the association between childhood problems and lifetime prevalence and costs of antidepressant medication by age 24 years. Method: A total of 5,547 subjects from a nation-wide birth cohort were linked to the National Prescription Register. Information about parent- and teacher-reported conduct, hyperkinetic and emotional symptoms, and self-reported depressive symptoms was gathered at age 8 years. The main outcome measure was national register-based lifetime information about purchases of antidepressants between ages 8 and 24 years. In addition, antidepressant costs were analyzed using a Heckman maximum likelihood model. Results: In all, 8.8% of males and 13.8% of females had used antidepressants between age 13 and 24 years. Among males, conduct problems independently predicted later antidepressant use. In both genders, self-reported depressive symptoms and living in other than a family with two biological parent at age 8 years independently predicted later antidepressant use. Significant gender interactions were found for conduct and hyperkinetic problems, indicating that more males who had these problems at age 8 have used antidepressants compared with females with the same problems. Conclusions: Childhood psychopathology predicts use of antidepressants, but the type of childhood psychopathology predicting antidepressant use is different among males and females. (Contains 4 tables and 1 figure.)