• We performed a three-night polysomnographic study of 54 rigorously assessed, drug free, prepubertal children who fit unmodified Research Diagnostic Criteria for major depressive disorder, and two groups of nondepressed controls (25 with emotional disorders and 11 who were normal). The groups did not differ polysomnographically, even though a high proportion of depressives and neurotics reported sleep disturbance in structured interviews. Sleep stage data do not appear to differentiate children with prepubertal major depressive disorders from nondepressed neurotic or normal children. Other psychobiologic findings in prepubertal depressives together with marked age effects on polysomnographic correlates of adult major depressive disorders suggest the hypothesis that polysomnographic abnormalities in adult major depressives are secondary to an interaction between depression and age.