Burden-of-illness assessments require incidence and prevalence assumptions. For psychotic disorders, there are few large-scale incidence and prevalence assessments providing rates for age, gender, or socio-economic subpopulations. Because outpatient treatment of psychotic conditions is increasing, overall incidence and prevalence may be estimated from outpatient pharmaceutical records. OBJECTIVE: This study estimates incidence and prevalence of antipsychotic utilization among age, gender, and payer combinations using outpatient pharmaceutical records. METHODS: Antipsychotic prescriptions dispensed between July 1995 and June 1997 to 4,628,528 eligible subjects were extracted from a large U.S database. Calculated ratios compared utilizers (eligible subjects with a prescription between July 1996 and June 1997) and initiators (utilizers without a prescription between July 1995 and June 1996) to eligible subjects for various combinations of age, gender, and payer types—Medicaid, non-Medicaid, indemnity, or “self-insured”. Chi-square tests were used to assess differences between those receiving and not receiving antipsychotic agents. RESULTS: The utilizer ratio was 5.0 per 1,000, although significantly (p = 0.002) lower (by 1.1 per 1000) for males. The initiator ratio was 1.6 per 1,000, again significantly (p <0.001) lower (by 0.1 per 1000) for males. Differences were not consistent—higher for younger males and lower for older males than for similarly aged females. Ratios were significantly (p ? 0.001) higher (by ten times) for Medicaid recipients. CONCLUSION: Outpatient ratios were within overall prevalence and incidence rates for schizophrenia or other psychotic conditions. Integrating these ratios into a burden-of-illness analysis should consider the proportions of non-psychotic disorders, patients treated for psychotic conditions without medication, and patients treated solely on an inpatient basis.