Objective: To identify childhood psychiatric symptoms as antecedents of cigarette smoking at age 18. Methods: In 1989, a general population sample of 2946 8-year-old boys born in Finland in 1981 was assessed using the Rutter's parent and teacher questionnaires, and the Child Depression Inventory. This birth cohort was followed up in 1999, when the subjects reported for their obligatory military service at age 18. Information about cigarette smoking frequency was obtained from 78% (n = 2307) of the boys attending the study in 1989. Results: Childhood hyperactivity and self-reported depressive symptoms correlated with moderate daily (1-10 cigarettes), and heavy daily (greater than 10 cigarettes) smoking at age 18. Conduct problems correlated with heavy daily smoking. A high level of childhood depressive symptoms, particularly in conjunction with a low educational level of the father, increased the risk of daily smoking. Emotional problems decreased the risk of smoking at age 18. In general, teacher reports had a better predictive power than parent reports for subsequent smoking. Conclusion: Future developmental studies with special focus on interaction between individual and environmental factors are warranted to reveal the mechanisms underlying the association between childhood psychopathology and adult smoking. In particular, the associations between childhood depression and future smoking need more clarification.