Eleven healthy males were studied twice. On one occasion (control, C), they slept (night 1) and then underwent a battery of tests at 4 h intervals from 06:00 day 1 to 02:00 day 2; then, after a normal sleep (night 2), they were tested from 10:00 to 22:00 on day 2. On the second occasion (sleep deprivation, SD), the subjects remained awake during night 1. Each battery of tests consisted of measurements of tympanic membrane temperature, profile of mood states (POMS), muscle strength, self-chosen work rate (SCWR), perceived exertion, and heart rate (HR) while exercising on a stationary cycle ergometer. Subjects also kept a diary of their activities during the two days and answered a questionnaire about their habitual physical activity. Results showed a significant negative effect of sleep deprivation on most mood states on day 1, but no effect on the other variables. By day 2, mood had tended to recover, though muscle strength tended to be worse in both control and sleep-deprivation experiments. There was also a more general tendency for negative effects to be present at the end of day 1 (02:00) or at the beginning of day 2 (10:00). There was limited support for the view that subjects who were habitually more active showed less negative effects after sleep deprivation and responded less adversely to the poor sleep achieved on the university premises (night 2). These results stress the considerable interindividual variation in the responses to sleep loss and, therefore, the difficulty associated with giving general advice to individuals about work or training capability after sleep loss.