The combined effect of midazolam (Ro 21-3981, Dormicum) and sleep deprivation on day-time sleep propensity was investigated in young, healthy adults. The oral administration of midazolam (15 mg) or placebo at bedtime was followed either by a sleep period of 7 or 4 h, or by no sleep at all. Sleep propensity was recorded at 2-h intervals throughout the following day by the multiple sleep latency test and self-ratings of tiredness. Performance was assessed in the morning and afternoon by a psychomotor test. Partial and total sleep deprivation caused a marked and significant increase of day-time sleep propensity. However, there was no significant difference between the midazolam and placebo condition after 7 or 4 h of sleep. After total sleep deprivation, sleep propensity was higher after placebo than after midazolam. Neither the treatment nor sleep duration had a significant effect on performance. The results show that a bedtime dose of 15 mg midazolam followed by normal or restricted sleep does not significantly affect day-time alertness.