The objective of this study was to monitor the long-term effect of clozapine administered to Parkinson's disease (PD) patients with psychosis. Confusion, visual hallucinations, and psychosis are major dose-limiting factors for long-term dopaminergic management of PD. Classic neuroleptic agents exacerbate the motor symptoms of the disease. For this reason, the introduction of atypical antipsychotic drugs has been a major advancement for the management of psychosis in patients with PD. Of them, clozapine is one of the most effective. Thirty-two patients (mean age, 73 years; mean disease duration, 12.2 years) with PD and psychosis (DSM-IV), 14 of them with dementia (DSM-IV), were followed for 5 years with periodic clinical evaluation, Mini Mental State Examination (MMSE), and Parkinsonian Psychosis Rating Scale (PPRS) administered before and following the study (at least once in 6 months). Periodic blood count was performed for tracking neutropenia. Nineteen patients (8 with dementia) have continued to receive clozapine (mean daily dose, 50 mg). Thirteen patients stopped medication: 9 because symptoms improved and did not return after weaning off clozapine; 3 patients because of somnolence; and 1 because of personal reasons. The average duration of treatment in those in whom medication was stopped was 8.5 months (range, 1–24 months). No correlation was found between age, sex, duration, and severity of disease (Yahr scoring), the presence of dementia, and the response to clozapine. Also, the PPRS scoring did not influence clozapine response. No case of neutropenia was found. According to the experience accumulated and the results of the present study, the authors believe clozapine is the best therapeutic choice currently available for the management of psychosis in patients with PD.