In a previous publication, prepubertal depressives were found not to exhibit a shortened REM period latency during active illness when compared with pathological and normal control groups. It was hypothesized that this might be due to a nonselective inhibition of REM sleep by slow-wave sleep (SWS), which is quite intense among prepubertal children, especially during the first 2 h of the sleep period. A number of polysomnographic signs normally associated with REM sleep were observed to occur periodically, prior to the beginning of the first REM period. It was thought that these signs might be indicative of "minor escapes" of REM sleep activity from the inhibitory influence of SWS. It was further hypothesized that differences among the experimental groups in the occurrence (timing and amount per minute) of these signs support the idea of an inhibitory effect of SWS on REM; and thus offer an explanation for the apparent dissociation between the depressive disorder among prepubertal children and a shortened REM period latency. The hypotheses were not supported by the results presented here indicating that the above dissociation is probably not due to inhibitory mechanisms of SWS on REM sleep.