Drug-induced improvement of depression may be mediated by changes in sleep physiology. In earlier studies on sleep EEG changes during treatment with antidepressants in depressed patients it could not be excluded that sleep disruptions and changes in the amount and distribution of REM sleep play a role in the changes in the sleep EEG. Therefore knowledge of the effects of antidepressants on the sleep EEG in healthy subjects with non-disturbed baseline sleep is necessary. In a three-way cross-over study in 12 healthy volunteers two single doses of Org 4428 (a highly specific noradrenaline reuptake inhibitor), 25 and 100 mg, were compared with placebo. Sleep EEGs were visually analysed and EEG power of non-REM sleep was measured. The results indicate that sole noradrenaline reuptake inhibiting activity is a potent mechanism to affect sleep polygraphic variables in an antidepressant-like way, i.e. REM sleep suppression and lengthening of REM latency. Despite the increase in the duration of non-REM sleep, i.e. stage 2, no significant changes in EEG power in the range 1–15 Hz were found. Therefore, the acute REM sleep suppression of Org 4428 did not result in a simultaneous reduction of EEG power during non-REM sleep. To date, these and earlier results indicate that most drugs with antidepressant properties affect REM sleep variables consistently, whereas their effect on both sleep polygraphic and EEG power variables in non-REM sleep is unpredictable.