Complaints of insomnia are common in patients with Parkinson's disease (PD). Melatonin has been shown to improve insomnia in some populations but has not been well studied in patients with PD. The primary aims of this pilot study were to assess endogenous melatonin secretion in PD patients by determining the time of dim-light melatonin onset (DLMO), and to compare the effect of exogenous melatonin (5 mg) therapy with placebo on nocturnal sleep in patients who complained of insomnia. A double-blind, placebo-controlled, cross-over trial was employed. Subjects (n=8) with PD and no evidence of depression, cognitive impairment, or primary sleep disorders participated in the 4-week protocol. During a 1-week treatment period, subjects took melatonin (5 mg) or placebo capsules (administered in random order) 30 minutes before bedtime, with a 1-week washout between treatments. DLMO was determined by RIA of blood samples. Nocturnal sleep was assessed by actigraphy. Subjective sleep quality was assessed with daily diaries and a weekly questionnaire. The mean DLMO was 21:05. During the melatonin treatment week there was a nonsignificant decrease in nocturnal waketime (20 minutes) and an increase in sleep efficiency (3%). Six subjects (75%) reported that they slept better during the melatonin treatment week. Results of this pilot study do not indicate that melatonin administration improves objective sleep in PD patients who complain of insomnia.