We gave 24 healthy elderly volunteers with a perceived sleep onset of at least 30 minutes zolpidem 5 mg, zolpidem 10 mg, or placebo for 7 days in a double-blind, three-way, crossover study. The morning after nocturnal dosing, psychomotor performance and cognitive ability were measured using tests which are sensitive to the residual effects of hypnotics and to the effects of drugs on various indicators of sleep quality. The tests were: Choice Reaction Time; Tracking; Critical Flicker Fusion Threshold; Memory Scanning; Word Recognition; the Leeds Sleep Evaluation Questionnaire and Line Analogue Rating Scales. Zolpidem produced a subjective improvement in sleep but did not impair performance the following day. Furthermore, during repeated administration, there was no tolerance to the effects of sleep latency and quality of sleep, nor adverse effects on task performance.