We compared twins with singletons in the National Epidemiological Child Psychiatric Study, which included 122 twins and 5455 singletons, born in 1981 and selected at random. Behavioural and emotional symptoms were assessed in 1989 on the basis of questionnaires filled in by the parents (Rutter Parent Questionnaire) (RA2), teachers (Rutter Teacher Questionnaire) (RB2) and the children themselves (Children's Depression Inventory) (CDI). Parents' reported proportions of probable behavioural/emotional disorders did not differ between the twin and singleton girls, but among the twin boys there was a nonsignificant trend of being more often probably disturbed. Twins were reported to be less disturbed than singletons according to the teachers' assessments. No difference was found between twins and singletons in their self-reports. When analysing parents' reported values of various sum scores, the twin boys obtained slightly higher scores than singletons, while twin girls scored significantly lower on total and emotional disturbances. Twin boys obtained lower mean scores than singletons for probable disorder in the teachers' evaluations, the differences arising mostly in the emotional area. The same type of trend, however nonsignificant, was found among the teachers' evaluations of girls. No significant difference was found in the mean scores for hyperactivity. This large population-based sample suggests that twins may have a lower rate of behavioural problems in childhood than singletons, a finding that has to be taken into account in behavioural genetic studies.