OBJECTIVE: To quantify differences in resource utilization and healthcare cost between olanzapine and risperidone when used as augmentation agents in patients with depression. METHODS: We conducted a retrospective analysis using the California Medicaid database (Medi-Cal 1996-2000). Adult patients aged 18 to 65 with 1 year of continuous enrollment and one or more medical claims for depression were included in the analysis. Patients with schizophrenia or psychotic depression were excluded. Treatment groups consisted of patients who received augmentation with risperidone or olanzapine after completing at least four weeks of antidepressant treatment. Cost comparisons were made using the amount paid by Medi-Cal to providers over the six-month period following the initiation of antidepressant treatment. RESULTS: Risperidone (N=105) and olanzapine (N=130) groups were similar with respect to length of treatment, use of antidepressants, and occurrence and number of outpatient and inpatient visits. Mean antipsychotic costs per month were significantly lower for risperidone ($154.31 vs. $258.13, p<0.0001). A significantly higher proportion of risperidone subjects (62.9% vs. 39.2% for olanzapine) had total mental healthcare costs that were below the median (p=0.0003). CONCLUSIONS: Relative to olanzapine, augmentation with risperidone was associated with significantly lower antipsychotic and total mental health costs.