A novel approach to investigate the relationship between depression and changes in sleep-wake regulatory mechanisms used the monoamine oxidase inhibitor (MAOI) phenelzine that is known to suppress rapid-eye-movement (REM) sleep. Sleep architecture and EEG topography during wakefulness and sleep were studied in eight depressed patients before and after five weeks of treatment with phenelzine (30-90 mg/day), which induced a significant alleviation of depressive symptoms. Theta power (4.75-7.5 Hz) during a 5-min wake EEG prior to sleep increased two-fold during administration of phenelzine. REM sleep was almost completely eliminated. This latter effect was compensated by increased duration of stage 2, whereas total sleep time was not shortened. In non-REM sleep (stages 2, 3, and 4), treatment slightly reduced EEG power between 2. Hz band increased. Activity in the delta band Mood and the regulation of sleep are closely related. This notion is supported by the presence of frequent sleep abnormalities in depressed patients, which may affect both rapid-eye-movement (REM) sleep and nonrapid-eye-movement (non-REM) sleep (Benca et al. 1992). A tight association between sleep and mood is further suggested by the beneficial action of total and partial sleep deprivation, as well as phase advance therapies of sleep in many patients with depression (Gillin and Borbély 1985; Wirz-Justice and Van den Hoofdakker 1999). Different theories have been proposed to account for the antidepressant effects of sleep manipulations. They include prolonged major suppression of REM sleep (Vogel 1975;Vogel et al. 1990), elevation of sleep propensity (Borbély and Wirz-Justice 1982), and reduction of non-REM sleep intensity (Beersma and Van den Hoofdakker 1992).The timing of sleep and wakefulness and the structure of sleep are regulated by the interaction of a homeostatic, sleep-wake-dependent process S and the circadian pacemaker located in the hypothalamus et al. 1981;Daan et al. 1984;Dijk et al. 1987;Achermann et al. 1993). Recent studies have indicated that an EEG correlate of sleep propensity can also be measured during wakefulness. Power in the theta band ( ف 5-8 Hz) of the wake EEG increases during sleep deprivation (Cajochen et al. 1995;Aeschbach et al. 1997;Finelli et al. 2000). The time constant of this increase is similar to that obtained for the wake-dependent rise of delta power as measured in non-REM sleep. Furthermore, topographical analyses of the EEG during prolonged wakefulness have shown that the increase of theta activity in waking and delta activity ( ف 1-4 Hz) in non-REM sleep are largest in frontal EEG derivations and positively correlated (Werth et al. 1998;Cajochen et al. 1999aCajochen et al. , 1999bFinelli et al. 2000). These findings suggest that sleep regulation may exhibit local features. In other words, frontal parts of the cortex may be particularly susceptible to sleep loss and reflect the homeostatic process of sleep regulation more sensitively than other cortical regions. The monoamine oxidase inhibitor...