Previous research has shown that caffeine and a < 15-min nap effectively and separately reduce sleepiness in drivers for 1 hr. In the present study, we examined in 12 sleepy individuals the treatments combined, taken during a 30-min break, prior to a longer (2 hr) continuous monotonous afternoon drive in a car simulator. Nonnap comparisons were 200 mg caffeine only and placebo. For placebo, driving incidents, subjective and electroencephalographic measures of sleepiness all reflected a mid-afternoon peak. This peak was significantly reduced by caffeine and eliminated by the combined treatment, which reduced incidents to 9% of placebo levels versus 34% of placebo levels for caffeine alone. Naps comprising "nonsleep dozing" were still effective.