Patients with sleep apnea/hypopnea syndrome (SAHS) suffer from impaired daytime function that correlates with hypoxemia and visible electroencephalographic (EEG) arousals. However, not all breathing irregularities during sleep terminate with visible EEG arousal. We hypothesized that sleep disturbance without visible EEG change may impair daytime function. Twelve normal subjects spent two pairs of 2 nights each in the laboratory. The first night of each pair was for acclimatization. On the second night, subjects either slept undisturbed or had sleep fragmented every minute to cause a transient increase in arterial blood pressure or increase in heart rate without visible EEG arousal. We tested daytime function after each study night. We presented 253 +/- 23 tones (mean +/- SD), 79 +/- 7% of which did not cause visible EEG arousals. Fragmentation did not alter total sleep time (undisturbed: 419 +/- 27 min; fragmented: 414 +/- 32 min; p = 0.5) or arousal frequency (undisturbed: 22 +/- 4/h; fragmented: 25 +/- 6/h; p = 0.4). Fragmentation reduced slow-wave sleep (undisturbed: 24 +/- 5%; fragmented: 20 +/- 4%; p < 0.01), mean sleep onset latency on the multiple sleep latency test (MSLT) (undisturbed: 8.0 +/- 3.1; fragmented: 6.2 +/- 2.1 min; p = 0.01) and the maintenance of wakefulness test (MWT) (undisturbed: 29.0 +/- 10.0 min; fragmented 25.7 +/- 9.7 min; p = 0.04). Fragmentation decreased hedonic tone at 7 A.M. (27 +/- 4, 25 +/- 6; p = 0.03). Nonvisible (autonomic) sleep fragmentation makes normal subjects sleepier and impairs their mood.