The effects on sleep of atenolol and clonidine were compared in 8 hypertensive men (mean age 46.9 years, range 16-56 years) without prior history of sleep disturbances. Polygraphic sleep recordings were performed at baseline (NO) and after a single oral dose of atenolol (100 mg) or clonidine (0.15 mg) at 6:00 pm at a 48-hour interval in a double-blind randomized crossover protocol. Both medications lowered arterial pressure to a similar extent. The subjective quality of sleep was judged satisfactory after both medications, the number of patients reporting dreams decreased from 5 (NO) to 1 after each treatment night. Total sleep time decreased slightly but not significantly after atenolol (440 +/- 63 min vs 474 +/- 47 min at baseline). Sleep latency was not affected after atenolol but significantly decreased after clonidine (16.9 +/- 21.6 vs 28.6 +/- 16.6 at baseline, p < 0.02). Although rapid-eye movement (REM) sleep time decreased after atenolol (71 +/- 30 min vs 95 +/- 30 min at baseline, p < 0.05), the percentage of REM sleep was unchanged (22 +/- 7% vs 23 +/- 5%). In contrast, clonidine strikingly reduced both REM sleep time (54 +/- 28 min vs 95 +/- 30 min at baseline, p < 0.002) and percentage of REM sleep (14 +/- 6% vs 23 +/- 5%, p < 0.0005). Thus atenolol tends to decrease total sleep time but does not affect the normal architecture of sleep, clonidine has a marked hypnotic effect, similar to that of some sedative medications and significantly reduces REM sleep.