Psychotic disorders in the elderly are frequent, of multiple etiologies, and little researched. With the advent of "atypical" neuroleptics, their role in treating elderly psychiatric patients needs to be investigated. Clozapine is widely used; however, its use is common in the elderly whose psychosis is a feature of neurological morbidity (Parkinson's disease, dementia, etc.), making it difficult to ascertain the safety, tolerability, and efficacy in psychiatric disorders in late life. The aim of the present review is to evaluate clozapine's effect in elderly psychiatric patients with no neurological comorbidity. A computerized literature search (MedLine 1966 to 1997) revealed 133 patients fulfilling said criteria. Fifteen patients had side effects and/or adverse events during treatment; nine of these were receiving a dosage greater than 100 mg clozapine daily. In 19 patients, treatment was discontinued, three due to noncompliance and 16 due to side effects. In seven patients, leukopenia/agranulocytosis was reported. The majority of side effects (27 of 34) and treatment discontinuations were within the first 90 days of treatment. Although efficacy is difficult to compare across studies because of differing methods of evaluation, the great majority of patients showed moderate to marked improvement of psychotic features. The reported effectiveness in patients able to continue treatment for extended periods is significant. Thus, clozapine at a relatively low mean dose (134 mg daily) seems to be safe, tolerated, and effective in elderly psychiatric patients. Agranulocytosis is more frequent than in younger adults and should be monitored carefully.