The hypnotic action of melatonin 5 mg p.o. was explored in 15 subjects with psychophysiological insomnia in a double-blind controlled self-report questionnaire study. Melatonin or placebo was taken at 20.00 hours for a 1-week period in random order. Effects on sleep and wakefulness were monitored by visual analogue scale and structured interview. Bedtime, sleep onset time, estimated total sleep and wake time, as well as self-rated sleep quality, were not altered by melatonin, and estimates of next-day function did not change. The period of melatonin treatment was retrospectively correctly identified by 8 of 15 subjects. Despite unchanged ratings of night sleep quality on the last night of each treatment, 7 of 15 subjects reported that sleep had subjectively improved to a minor extent in the week of active treatment. Side-effects attributed to melatonin included headache and an odd taste in the mouth. These data indicate that melatonin is probably of no clinical value in the management of psychophysiological insomnia.