Objective: This study examines total pharmacy cost and usage patterns of schizophrenic patients in acute mental health inpatient settings for three atypical antipsychotics risperidone, olanzapine, and quetiapine. Despite the readily available unit cost information for drugs, actual pharmacy costs may deviate significantly from 'labeled costs'. Recent research findings indicate the need for more robust evaluation of such pharmacy costs. Research design and methods: This study used data from non-randomized inpatient retrospective charts from three acute care inpatient mental health facilities. The final pooled sample included 327 patients, of which 120 received risperidone, 153 received olanzapine, and 54 received quetiapine. Medication cost was defined as the average wholesale price (AWP) as listed in the 2001 'Red Book'. Propensity scoring methodology and multinomial regression were employed to reduce treatment selection bias. Results: The observed mean daily antipsychotic drug doses were 4.45 mg (SD 2.44) for risperidone, 14.04 mg (SD 5.55) for olanzapine, and 350.33 mg (SD 228.24) for quetiapine. The corresponding mean daily drug costs were $7.66(SD $4.20) for risperidone, $8.11 (SD $5.29) for quetiapine and, $12.10 (SD $4.79) for olanzepine. Numbers adjusted for treatment selection bias show that the average daily total pharmacy cost of risperidone was $4.35 lower than olanzapine (p < 0.001) and $1.41 lower than quetiapine (p = 0.38). The adjusted average daily pharmacy cost of olanzapine was $4.02 higher than quetiapine (p < 0.001). After statistical adjustment there were no significant differences between study drugs in terms of length of stay or patient functioning. Conclusion: This study provides the first US comparison of medication utilization patterns and pharmacy costs for olanzapine, risperidone, and quetiapine administered in acute mental health care inpatient settings. While this study did not estimate the full economic value of the three antipsychotics in these inpatient settings, it demonstrated that the mean daily costs for risperidone were lower than the mean daily costs for olanzapine (p < 0.001) and quetiapine although the later difference was not statistically significant (p = 0.38)