Levodopa-induced psychotic symptoms frequently complicate the management of patients with Parkinson's disease (PD). We examined the efficacy and tolerability of a novel antipsychotic, remoxipride, in this population. This was a 7-week, open-label pilot evaluation of patients with moderate to severe PD and levodopa-induced psychotic symptoms of at least 2 months' duration. The patients were recruited at the Movement Disorders Clinic, The Toronto Hospital, a tertiary referral center. After 1 week of baseline observation, the patients received remoxipride, 25 mg, three times a day orally, with the dose increasing by 25 to 50 mg each week as tolerated. The outcome measures included the Brief Psychiatric Rating Scale (BPRS), the Clinical Global Impressions (CGI) scale, and the United Parkinson's Disease Rating Scale. Adverse symptoms were elicited by an open-ended questionnaire and a symptom checklist. Six men and three women aged 69.3 +/- 9 years received remoxipride 147 +/- 57 mg/day. Total BPRS score decreased modestly in eight of nine subjects, and there was a statistically significant improvement of mental status as indicated by the CGI scale score, which decreased from 3.8 +/- 0.4 at baseline to 2.4 +/- 1.3 at last rating (p < 0.05; Wilcoxon signed rank test). The motor performance deteriorated somewhat in two subjects, whereas the rest showed no appreciable change. The most common adverse effects included tremor, rigidity, akathisia, and hypersalivation. Remoxipride treatment reduced psychotic symptoms in eight of nine subjects while having no appreciable effect on the parkinsonian status of seven of nine subjects.