OBJECTIVES: To evaluate the direct health care costs associated with olanzapine, risperidone, and quetiapine monotherapy among patients diagnosed with bipolar disorder (ICD-9: 296.4x-296.8x). METHODS: Using a sample drawn from the NC Medicaid Claims database during August 2000 through January 2005. This study included patients with a bipolar-related diagnosis who were naïve to atypical antipsychotic treatment and were without a bipolar-related medical visit or hospitalization during 90 days prior to treatment initiation. Patients were followed for 12 months after initiation of atypical antipsychotic monotherapy (index drug). Costs of index drug, all bipolar-related medical care, and all health-related costs, both including and excluding index drug, were examined in the 12 month treatment period using Generalized Linear Model with Gamma Distribution and Log link. To account for potential confounds, the model included several covariates. RESULTS: A total of 838 continuously eligible patients met the inclusion criteria (393 olanzapine, 262 risperidone and 183 quetiapine). The costs of index drug for patients taking olanzapine were 43% (P<0.0001) and 19% higher (p<0.0001) than risperidone and quetiapine, respectively. In terms of total health-related cost there was no difference between patients treated with olanzapine and those treated with risperidone or quetiapine, including or excluding index drug. In terms of all bipolar-related medical care costs, the inclusion of index drug led to 15.2% (p<0.04) higher costs for patients receiving olanzapine compared to risperidone, primarily due to the higher acquisition cost of olanzapine. CONCLUSIONS: Despite significantly higher acquisition costs of olanzapine when used as mono-therapy for the treatment of bipolar disorder, total health-related costs with and without index drug were similar for olanzapine, risperidone and quetiapine. Bipolar-related medical costs excluding index drug were also similar for olanzapine, risperidone, and quetiapine treatment. However, the inclusion index drug costs resulted in higher bipolar-related medical costs for patients receiving olanzapine compared to risperidone.