Eight young adults underwent 1 night of auditory sleep fragmentation followed by four naps of the multiple sleep latency test and performance testing the next day. A latin-square design was used to compare results with baseline. Efforts were made to eliminate effects of learning on repeated performance tests. A mean of 303 arousals, lasting a mean of 11 seconds, disrupted nocturnal sleep. This sleep fragmentation was induced to mimic as closely as possible the nocturnal sleep disruption seen in subjects with upper airway resistance syndrome. There was a significant disruption of nocturnal sleep architecture with a significant overall decrease in slow-wave sleep (SWS) and a significant but more moderate decrease in rapid eye movement (REM) sleep during the fragmented night. The most interesting finding related to analysis by thirds of the night, which indicated an important increase over time in arousal threshold during SWS followed by REM sleep. This threshold increase was associated with a parallel increase in dB(A) levels needed to induce an arousal. Stages 1 and 2 nonrapid eye movement (NREM) sleep were less affected by the stimulation, but the amount of stage 1 NREM sleep decreased from the beginning to the end of the night, again indicating an increase in pressure to sleep. Following 1 night of sleep fragmentation, subjects had significantly shorter sleep latencies on the multiple sleep latency test for naps 2, 3 and 4. There was a significant relationship between percent nocturnal SWS and mean sleep latencies. The selected performance tests were not affected by 1 night of sleep fragmentation, despite the obvious sleepiness.