In a study of 1600 Dutch adolescent twin pairs we found that 59% of the inter-individual variation in smoking behaviour could be attributed to shared environmental influences and 31% to genetic factors. The magnitude of the genetic and environmental effects did not differ between boys and girls. However, environmental effects shared by male twins and environmental effects shared by female twins were imperfectly correlated in twins from opposite-sex pairs, indicating that different environmental factors influence smoking in adolescent boys and girls. In the parents of these twins, the correlation between husband and wife for 'currently smoking' (r = 0.43) was larger than for 'ever smoked' (r = 0.18). There was no evidence that smoking of parents (at present or in the past) encouraged smoking in their offspring. Resemblance between parents and offspring was significant but rather low and could be accounted for completely by their genetic relatedness. Moreover, the association between 'currently smoking' in the parents and smoking behaviour in their children was not larger than the association between 'ever smoking' in parents and smoking in their children.