Ten patients with Parkinson's disease and levodopa induced dyskinesias (LIDs) took part in this randomised, placebo controlled, double blind, crossover trial to assess the efficacy and tolerability of high dose oral naltrexone for LIDs in Parkinson's disease. Patients received naltrexone (5 mg/kg/day) or placebo for 2.5 weeks with 1 week wash out in between. Dyskinesias and motor function were assessed with a levodopa challenge, unified Parkinson's disease rating scale (UPDRS), the unified dyskinesia rating scale (UDRS), and patient diaries. Eight patients completed the trial. There was a small reduction in LIDs measured by patient diaries with naltrexone (20.5 (SD 24.9)%) compared with placebo (-4.1 (SD 22.6)%), p<0.05, although no difference was found by other subjective or objective measures. Naltrexone was well tolerated and caused no significant differences in UPDRS motor scores or off time. This study suggests that short term therapy with high dose naltrexone (250-350 mg/day) has no or minimal effect on reducing LIDs in Parkinson's disease.