OBJECTIVES: Recent research has shown that antidepressants may improve non-cognitive symptoms in patients with Alzheimer's disease (AD). The economic benefits of antidepressant use in this population are unknown. This study investigated whether antidepressants use by AD patients could delay the admission to long-term care (LTC) treatment and reduce total costs. METHODS: Patients diagnosed with AD from 1995-2001 were retrieved from a 20% sample of the California Medicaid (Medi-Cal) claims database. The index date for the antidepressant group was defined as the dispensing date of first antidepressant medication, and for the control group, it was defined as the date of initial AD diagnosis. Generalized linear models (GLM) with log link function and gamma distribution were applied to compare the first year total post-index medical (non-pharmacy) costs and the total health care costs between the groups, controlling for demographics, comorbidities, and prior utilization. Cox proportional hazard regression was used to model the time to first LTC admission since index date, adjusting for the same set of covariates. RESULTS: In total, 3515 AD patients were identified, with 1645 in the antidepressant group and 1870 in the control. AD patients with antidepressants were younger (77 vs. 80 years, p<0.0001), had higher proportion of Caucasians (57% vs. 53%, p<0.0001), and had a similar comorbidity profile as the control group. Adjusted for potential confounders, the GLM estimates showed that using antidepressants was related to 40% and 36% lower post-index total medical costs and total health care costs (p<0.0001 for both), respectively. In the Cox regression, antidepressant group had 34% lower risk of admission to LTC than the control group (P<0.0001). CONCLUSIONS: In this study, antidepressant use in AD patients associated a significant reduction in total costs and appeared to postpone long-term care admission.