OBJECTIVES To examine utilization patterns and hypoglycemia in patients with type-2 diabetes mellitus (T2DM) receiving concomitant exenatide and long-acting insulin therapy over 6 months. METHODS A retrospective analysis was performed using claims data (January 1, 2004 to March 31, 2008) from a large US prescription database. Adult patients with T2DM and no claim for exenatide (index drug) in the previous 12 months were included. At least 12 months pre-index and ≥6 months post-index continuous eligibility were required. Concomitant long-acting insulin use was defined as a claim for NPH, insulin glargine or insulin detemir 100 days pre-index to 15 days post-index. Therapy discontinuation was defined as a >60 day gap between prescription claims. Hypoglycemic events were identified by ICD-9 codes 250.8, 251.0, 251.1, and 251.2. Hypoglycemia generating a claim using these codes is generally severe. RESULTS Concomitant exenatide and long-acting insulin use was identified in 2082 patients. Of these, 38% discontinued both therapies, 15% discontinued exenatide but remained on long-acting insulin, 24% discontinued long-acting insulin but remained on exenatide and 23% continued both therapies. There were no differences in age, gender or Deyo-Charleson comorbidity index score among cohorts. Patients who discontinued exenatide and remained on long acting insulin had a higher mean rate of hypoglycemia (0.10±1.01 events/patient/6-months), compared to patients who discontinued long-acting insulin but remained on exenatide (0.02±0.20), discontinued both therapies (0.04±0.37) or continued both therapies (0.03±0.26; p<0.0001 across groups). Insufficient pre- and post-A1C values were available for an analysis of this endpoint. CONCLUSIONS Patients remaining on concomitant exenatide and long-acting insulin did not experience more hypoglycemic events than patients who remained on either drug alone. However, patients discontinuing exenatide but remaining on long-acting insulin experienced significantly more hypoglycemic events. The increase in hypoglycemic events may have been due to an increase in basal insulin dose and/or the addition of other insulin therapy.