We studied 38 patients with Parkinson's disease (PD) without dementia (mean age, 67.3±4.8 years; 15 males, 23 females) with complaints on sleep disorders. Quality of sleep was assessed with the Parkinson's disease sleep scale (PDSS) and the Epworth Sleepiness Scale (ESS) as well as with overnight polysomnographic (PSG) study at baseline and at the end of the trial. The effectiveness of sleep was estimated as TST/TIB x100% (TIB - habitual time in bed, TST - habitual total sleeping time). REM latency (LREM), periodic limb movements (PLM) (total number and sleep index) were measured as well. All patients underwent neuropsychological testing using MMSE, five-word test, digit span and the Hamilton scale (HAM-D). Patients were allocated to 2 groups. Group 1 (n=20) received melatonin in addition to the previous dopaminergic treatment in dose 3 mg 30 minutes before bedtime for 6 weeks, group 2 (n=18) received clonazepam 2 mg at night (with gradual titration over 4 weeks from 0,5 mg). Compared to baseline, melatonin and clonazepam reduced sleep disorders in patients: PDSS scores from 89.9±8.9 to 129.5±9.4 (p=0.0001) and from 91.0±8.7 to 110.1±12.4 scores (p=0.03), respectively. However, the daytime sleepiness (ESS) was significantly increased (from 3.8±1.2 to 7.3±2.2 scores; p=0.0002) in the clonazepam group. In the melatonin group, ESS scores were 4.1±1.4 before treatment and 4.7±1.4 after treatment (p=0.06). Patients treated with melatonin had better scores on the MMSE (p=0.00009), five-word test (p=0.009), Hamilton scale (p=0.00009) at the end of the study period as compared with the clonazepam group. Changes in total point scores on the PSG at the end of week 6, as compared with the beginning of the trial, were in favor of the group treated with melatonin, with significant changes in the LS (p=0.004), total sleep time/time in bad (TST/TIB) (p=0,001) sections. The number of REM sleep epochs remained lower in patients treated with clonazepam (p=0.0001). The data suggest high treatment efficacy of melatonin in the treatment of sleep disorders in Parkinson's disease without dementia.