OBJECTIVE: To examine antidepressant medication utilization and adherence of managed care patients enrolled in a medication adherence program. METHODS: This was an aggregate, retrospective cohort analysis of pharmacy and eligibility claims databases from four health plans. The study cohorts included antidepressant new starts (no antidepressant in the previous 4 months) 18 years or older between September 1, 2001 and July 31, 2002 who were continuously enrolled during the study period and were nonadherent with therapy. The nonadherent criteria for 3 plans was > 10 days late in filling medication and was > 20 days late for 1 plan. The intervention cohort consisted of patients enrolled in the program whose physicians were notified of nonadherence in filling antidepressant medications. The control cohort included nonadherent patients not enrolled in the program. A t-test was used to evaluate differences in medication utilization. A multivariate linear regression analysis was performed to estimate the association of the intervention with medication adherence (defined as medication possession ratio [MPR]) when adjusted for relevant covariates. RESULTS: The control (N = 13,152) and intervention (N = 2,296) cohorts were comparable in age and gender; however, the control had a higher mean copay per patient per prescription ($16.93 vs. $12.33; p <0.0001). Six-month evaluation of antidepressant pharmacy claims showed the intervened patients had a higher average number of prescriptions (4.5 vs. 4.1), quantity dispensed (213 vs. 164) and days coverage (140 vs. 124); p <0.0001. The intervention cohort lowered the mean number of gap days between medication fills (32 vs. 54; p <0.0001) and improved mean MPR (81% vs. 67%; p <0.0001). Multivariate analysis demonstrated MPR was associated with enrollment in the program (p <0.0001), copay (p = 0.0131), therapy initiation with selective serotonin inhibitors (p <0.0001) and > 20 day nonadherent criteria (p <0.0001). CONCLUSIONS: Antidepressant medication utilization and adherence can improve with physician notification of nonadherent patients. The nonadherent criteria of > 20 days negatively impacted the aggregate MPR.