OBJECTIVES: To measure the acute-phase persistence and cost of antidepressant therapy among commercially insured patients having diabetes.  METHODS: We conducted a retrospective study among patients with diabetes receiving health insurance through Blue Cross & Blue Shield of RI between July 1, 2008 - December 31, 2009. New users of antidepressants were defined as having no history of an antidepressant dispensing for at least 90 days preceding an incident antidepressant prescription.  Patients that continued their medication for at least for 90 days were classified as persistent. We compared persistence rates according to the therapeutic class of antidepressant prescribed, and by users of brand versus generic products. A multivariate logistic regression model was developed to predict non-persistence with antidepressant therapy by antidepressant type (SSRI/non-SSRI), and by use of generic versus brand products. The average total prescription cost for the 90 day period was calculated and stratified by use of brand or generic antidepressants. RESULTS: We identified 743 patients with diabetes who were newly prescribed an antidepressant; a majority were prescribed SSRI antidepressants (52.5%), while 26% were prescribed a brand name antidepressant, and 432 (58%) patients were classified as non-persistent. Users of non-SSRI antidepressants were more likely to fail to persist (OR 1.57, 95% CI 1.16-2.12). For non-persistent patients, the average cost of therapy was $126.70 among users of brand antidepressants, while the average cost among users of generics was $17.10. Among persistent patients, the average cost among users of brand antidepressants was $461.4, while the average cost among users of generics was $70.80. CONCLUSIONS: In this population of commercially-insured patients having diabetes, acute-phase persistence with antidepressant therapy was higher among patients prescribed SSRI medications, while treatment costs differed substantially among those prescribed brand versus generic medications.