The separate and combined effects of prolonged wakefulness and alcohol were compared on measures of subjective sleepiness, simulated driving performance and drivers' ability to judge impairment. Twenty-two males aged between 19 and 35 years were tested on four occasions. Subjects drove for 30 min on a simulated driving task under conditions determined by the factorial combination of 16 and 20 h of wakefulness and blood alcohol concentrations of 0.00 and 0.08%. The simulated driving session took place 30 min postingestion; subjects in the two alcohol conditions participated in a second 30-min driving session 90-min postingestion. Subjects made simultaneous ratings of their impairment while driving and retrospective ratings at the end of each test session. Subjective sleepiness measures were completed before and after each driving session. The combination of 20 h of prolonged wakefulness and alcohol produced significantly lower ratings of subjective sleepiness and driving performance that was worse, but not significantly so, than would be expected from the additive effects of each condition alone. Driving performance was always worse in the second driving session, during the elimination phase of alcohol metabolism, despite blood alcohol concentrations being lower than during the first driving session. There was a modest association between perceived and actual impairments in driving performance following prolonged wakefulness and alcohol. The findings suggest that the combination of prolonged wakefulness and alcohol consumption produced greater decrements in simulated driving performance than each condition alone and that drivers have only a modest ability to appreciate the magnitude of their impairment.