OBJECTIVE: Examine the mean difference in hospitalization and emergency room (ER) visit rates pre and post-initiation on olanzapine, quetiapine or risperidone. METHODS: Retrospective analysis of a commercial claims database capturing inpatient, outpatient and prescription drug services. Individuals having a mental illness diagnosis, and initiated with olanzapine (N=8730), quetiapine (N=5709) or risperidone (N=9339) between July 1, 1998 and July 2, 2002 were identified. Mean differences in the hospitalization or ER visit rates between the pre- and post- six month periods across the three-treatments were examined using ANCOVA, controlling for age, gender, region, and type of mental illness diagnosis. RESULTS: Individuals initiated with olanzapine or risperidone were found to have a significantly higher difference in the hospitalization (2.19%, p<0.0001; 1.72%, p<0.0001) or ER visit rates (3.80, p<0.0001; 4.60%, p<0.0001) post initiation of medication compared to the six months prior to initiation of medication. In contrast, individuals who were initiated with quetiapine had a significantly lower difference in the hospitalization (-4.37%, p<0.0001) or ER visit rates (-2.89%, p<0.0001) post-initiation of quetiapine compared to the six months prior-initiation of quetiapine. CONCLUSION: Quetiapine, unlike olanzapine or risperidone, may be associated with fewer hospitalizations and ER visits after medication initiation. These results may be suggestive of a more favorable side effect profile and/or better compliance with Quetiapine, and needs further investigation.