Recent evidence shows that the temporal alignment between the sleep-wake cycle and the circadian pacemaker affects self-assessment of mood in healthy subjects. Despite the differences in affective state between healthy subjects and patients with psychiatric disorders, these results have implications for analyzing diurnal variation of mood in unipolar and bipolar affective disorders and sleep disturbances in other major psychiatric conditions such as chronic schizophrenia. In a good proportion of patients with depression, mood often improves over the course of the day; an extension of waking often has an antidepressant effect. Sleep deprivation has been described as a treatment for depression for more than 30 years, and approximately 50% to 60% of patients with depression respond to this approach, especially those patients who report that their mood improves over the course of the day. The mechanisms by which sleep deprivation exerts its antidepressant effects are still controversial, but a reduction in rapid eye movement sleep (REM sleep), sleep pressure and slow-wave sleep (SWS), or a circadian phase disturbance, have been proposed. Although several studies support each of these hypotheses, none is sufficient to explain all observations reported to date. Unfortunately, the disturbed sleep-wake cycle or behavioural activities of depressed patients often explain several of the abnormalities reported in the diurnal rhythms of these patients. Thus, protocols that specifically manipulate the sleep-wake cycle to unmask the expression of the endogenous circadian pacemaker are greatly needed. In chronic schizophrenia, significant disturbances in sleep continuity, REM sleep, and SWS have been consistently reported. These disturbances are different from those observed in depression, especially with regard to REM sleep. Circadian phase abnormalities in schizophrenic patients have also been reported. Future research is expected to clarify the nature of these abnormalities.