After the chronic administration of saline, rats were treated with nomifensine (0.1 or 1.0 mg/kg, twice a day, at light and dark onset) for 11 days. The sleep-wake activity was recorded for 24 h on the baseline day (saline), on nomifensine days 1, 5 and 11, and also on day 12, when saline was injected again (withdrawal day). Another group of rats was treated with saline throughout the experiment, without significant effect on the sleep-wake activity. The smaller dose of nomifensine increased non-REM sleep (NREMS) at the expense of wakefulness (W) in the light period. The effect persisted throughout the chronic treatment. A late increase in REM sleep (REMS) was noted on nomifensine days 5 and 11. Nomifensine failed to affect the sleep-wake activity in the dark period. On withdrawal, the baseline percentages of the vigilance states were recovered. As evaluated through spectral analysis of the EEG, the increase in NREMS was accompanied by an increase in slow wave activity. The higher dose of nomifensine elicited an increase in W and a reduction in both sleep states, followed by changes in W and NREMS in the opposite directions. These effects were evident in both the light and the dark periods of the day. Chronic treatment resulted in circadian variations in the effects. Withdrawal of the drug abolished the arousal reaction, but the late increase in NREMS persisted. The dose-dependent biphasic effects of nomifensine on sleep-wake activity can be explained by considering the proposed indirect dopamine and possibly noradrenaline agonist activity of the drug.(ABSTRACT TRUNCATED AT 250 WORDS)