A single dose of ethanol (0.60 g/kg of body weight) was administered to eight young healthy male subjects 35 minutes before bedtime. Compared to the average value of two baseline nights, subjective sleep and polysomnographically determined sleep parameters were not significantly affected. In the first 2 hours of sleep after ethanol intake, the combined value of wakefulness, stage 1, and movement time was reduced. In this interval, visually scored stage 4 sleep was increased, and electroencephalographic (EEG) power density in nonrapid-eye-movement (nonREM) sleep was enhanced in the lowest delta frequencies and reduced in the beta range. Computed for the entire sleep episode, power density in REM sleep was enhanced in some theta frequencies. In the sleep episode initiated 24 hours after ethanol intake, power density in nonREM and REM sleep was enhanced in delta and theta frequencies, and the subjectively perceived number of awakenings was reduced. The effects of ethanol on EEG power spectra during sleep differ from those published for benzodiazepine and nonbenzodiazepine hypnotics. This indicates that the effects of ethanol on the human sleep EEG are not mediated by the benzodiazepine receptor.