Patients with type 2 diabetes mellitus are routinely treated with combinations of glucose-lowering agents. The adverse event (AE) profile and effects on glycemic control have not been assessed for the glucagon-like peptide-1 receptor agonist exenatide once weekly in combination with a thiazolidinedione (TZD) with or without metformin. This study was conducted to examine the long-term safety profile and changes in glycemic control and weight for exenatide once weekly with TZD with or without metformin in patients with type 2 diabetes mellitus over 2 years. In this single-arm, open-label trial with treatment up to 104 or 117 weeks, patients received 2 mg exenatide once weekly while continuing treatment with a TZD with or without metformin. Patients were either exenatide-naïve before this study or had previously received exenatide twice daily, which was discontinued on initiating exenatide once weekly. Patients were on a stable dosage of TZD (rosiglitazone or pioglitazone) and, if applicable, metformin. Treatment-emergent AEs were defined as those first occurring or worsening post baseline. Descriptive statistics were used for absolute and change-from-baseline data, and a one-sample t test for within-group change in glycosylated hemoglobin (HbA 1c ). Of 134 patients in the intent-to-treat population (baseline mean [SD] HbA 1c ,7.2% [1.0%]), 44 were exenatide-naïve (baseline HbA 1c , 7.8% [1.0%]) and 90 switched from exenatide twice daily (baseline HbA 1c , 7.0% [0.8%]). Of intent-to-treat patients, 106 (79%) completed the final treatment visit (week 104 or week 117). The most common AEs were nausea (17% of patients) and injection-site nodule (12% of patients). Serious AEs were reported in 14% of patients and 5% withdrew because of a treatment-emergent AE. No identifiable pattern of serious AEs was observed. There were 4 reports of edema and no reports of heart failure. No major hypoglycemia was reported; minor hypoglycemia was reported in 4% of patients. Exenatide-naïve patients experienced mean (SE) HbA 1c reductions of −0.7% (0.2%) and weight reductions of −2.7 (0.8) kg, whereas patients with prior exposure to exenatide twice daily experienced a reduction of −0.4% (0.1%) in HbA 1c and no change in weight. Adverse events over 2 years were consistent with the reported safety profiles of exenatide once weekly and TZDs. Exenatide-naïve patients experienced improvements in HbA 1c and weight, while patients with the benefit of prior exenatide therapy experienced an additional reduction from baseline in HbA 1c and no additional change in weight after 2 years. ClinicalTrials.gov identifier: NCT00753896.