This study used administrative claims data to compare the relative risks for hospitalisation among patients with schizophrenia within a US Medicaid programme receiving atypical and typical antipsychotics. The newer atypical antipsychotics may be better tolerated among mentally ill patients receiving public assistance (Medicaid) who are less functional than other mentally ill populations. Risperidone, olanzapine, quetiapine and ziprasidone were compared with each other and to typical antipsychotics as a single category. Cox proportional hazard estimates, adjusted for differences in patient characteristics, showed numerically lower risks for each of the atypicals in comparison with the typicals, with that for quetiapine being statistically significant (HR: 0.672, p = 0.0413). There were no statistically significant differences among atypical pairs. This study provides evidence that risk for hospitalisation among Medicaid patients with schizophrenia may be lower with atypical antipsychotics, particularly quetiapine.