In a double-blind, parallel groups study, 60 healthy student volunteers (29 men and 31 women, aged 19-22 years) were randomly allocated to receive placebo, 100 or 200 mg modafinil. Two hours later, in the early evening, they completed an extensive cognitive battery. The 3 groups did not differ in self-ratings of sleepiness or tiredness before the testing session, and there were no treatment-associated changes in these or in mood ratings during the tests. Modafinil was without effect in several tests of reaction time and attention, but the 200-mg group was faster at simple color naming of dots and performed better than placebo in the Rapid Visual Information Processing test of sustained attention. Modafinil was without effect on spatial working memory, but the 100-mg group performed better in the backward part of the digit span test. Modafinil was without effect on verbal short-term memory (story recall), but 100 mg improved digit span forward, and both doses improved pattern recognition, although this was accompanied by a slowing of response latency in the 200-mg group. There were no significant effects of modafinil compared with placebo in tests of long-term memory, executive function, visuospatial and constructional ability, or category fluency. These results suggest that the benefits of modafinil are not clearly dose-related, and those from 100 mg are limited to the span of immediate verbal recall and short-term visual recognition memory, which is insufficient for it to be considered as a cognitive enhancer in non-sleep-deprived individuals.