Young male Naval volunteers were denied normal noclurnal sleep and maintained on a 60-min lreatment-160-min testing schedule during 40 consecutive hrs. Ten subjects bicycled, 20 subjects controlled EEG activity during bedrest, and 10 subjects napped. Eight measures of addition, auditory vigilance, mood, and oral temperature were obtained. The Bedrest group showed significant impairment on all eight measures, and thus, gave no support to lite forced-rest theory of sleep function. The Exercise group was worse than the Nap and Bedrest groups for all measures. In spite of fragmented, reduced sleep (about 3.7 hrs per 24 hrs), the Nap group had no impairment on six of the measures. The results suggest that exercise increases the impairment due to sleep loss, and naps reduce or remove this impairment. Bedrest is not a substitute for sleep.