Schizophrenia affects less than 1% of the US population, yet its treatment accounts for more than 2.5% of total healthcare expenditures, making it the most costly illness to treat in psychiatry. OBJECTIVES: The purpose of this study is to document the treatment experience of patients with schizophrenia in a managed care population. METHODS: This study utilized an administrative claims database from PacifiCare in California. Prescription and encounter claims from 1/1/95-9/1/99 in California, Texas, Oklahoma, Washington and Oregon were evaluated using an intent-to-treat analysis. Only adults were included, with ?6 prior months eligibility, no antipsychotic medication use in the 120 days before treatment start (washout period), continuous antipsychotic use for ?120 days, and ?365 days of eligibility following treatment start. RESULTS: Of the 4,388 PacifiCare members qualifying for this analysis, only 232 (5%) had a diagnosis of schizophrenia in the 180 days prior to the treatment start. In 4,321 (98%) of the antipsychotic treatment episodes, a single antipsychotic agent was initiated. Of those initiating antipsychotic mono-therapy, 3,286 (76%) began therapy on a conventional medication, while 1,035 (24%) started on an atypical medication. Only 67 patients initiated treatment using ?2 antipsychotics simultaneously. Time-to-event analysis of treatment discontinuation (mono-atypical vs. mono-conventional vs. multi-therapy) found that the probabilities of discontinuing therapy differed by year of prescription. In more current prescription years, patients initiating atypical therapy were more likely to remain on therapy than those starting on conventional agents. CONCLUSION: Atypical agents appear to provide improved performance in treatment continuation.