This was a placebo-controlled, double-blind randomized crossover study of long-term (5 weeks) administration of fluoxetine (20 mg/day) and dothiepin (75 mg/day for 1 week followed by 150 mg/day for 4 weeks) in 12 healthy male volunteers. Subjects were studied on day 10 and day 36 of treatment, with tests of nocturnal sleep, driving performance, continuous electroencephalogram (EEG), sleep during scheduled naps, computerized visual attention tasks, saccadic eye movement measurement and visual analogue ratings of mood. Both drugs had a marked suppressive effect on nocturnal rapid eye movement (REM) sleep; these effects were less at 36 days than at 10 days, and fluoxetine decreased and dothiepin increased REM in daytime naps. Sleep fragmentation after fluoxetine is similar to that reported in the literature. We found no sleep-promoting effects of dothiepin, in contrast to our previous single-dose study, and no subjective sleep effects of either drug. Subjects were less sleepy after both antidepressants than placebo at 5 weeks measured by sleep latencies and EEG. Saccadic eye movement measures were significantly faster after 5 weeks of fluoxetine than after 5 weeks of placebo. Reaction times to a peripheral stimulus during computerized tracking task were shorter after 10 days of dothiepin compared with placebo. Driving performance, visual attention and mood ratings showed no treatment effects. Subjective health reports during each 5 weeks of treatment were similar in number for the two drugs but showed a different profile of side-effects.