OBJECTIVE: This study was conducted to compare the efficacy and safety profiles of exenatide and insulin glargine therapy in patients with type 2 diabetes who had not achieved glucose control with metformin or sulfonylurea monotherapy. METHODS: This multinational, randomized, open-label, crossover noninferiority study compared the efficacy of exenatide 10 pg BID and insulin glargine QD (titrated targeting a fasting serum glucose (FSG) level < or =5.6 mmol/L) in patients with type 2 diabetes treated with a single oral antidiabetic agent. The study included two 16-week treatment periods. The primary a priori outcome variable was the change in glycosylated hemoglobin (HbA(lc)). Secondary outcomes included the proportion of patients achieving the American Diabetes Association (ADA) target HbA(lc) of < or =7% and the European Association for the Study of Diabetes target of < or =6.5%, the change in FSG, end-point values and change in the 7-point self-monitored glucose profile, and change in body weight. Adverse events were assessed based on standard laboratory tests and patient reports. RESULTS: One hundred thirty-eight patients were randomized to study treatment (52.9% female, 47.1% male; 79.7% white; mean [SEM] age, 54.9 [0.8] years; duration of diabetes, 7.4 [0.4] years; body mass index, 31.1 [0.4] kg/m(2); weight, 84.8 [1.4] kg) while continuing to receive metformin (55.1%) or a sulfonylurea (44.9%). The population had a baseline least squares (LS) mean (SEM) HbA(lc) of 8.95% (0.09%) and an LS mean FSG concentration of 12.0 (0.3) mmol/L. Both exenatide and titrated insulin glargine therapy were associated with similar significant changes from baseline in HbA(1c) (both, -1.36% [0.09%]; P < 0.001); the difference between groups was not statistically significant. The LS mean HbA(1c) at end point was above the ADA target with both treatments (exenatide, 7.57% [0.09%]; insulin glargine, 7.58% [0.09%]). Similar proportions of patients achieved an HbA(1c) < or =7% (37.5% and 39.8%, respectively; P = NS) or < or =6.5% (21.5% and 13.6%). Patients lost weight during exenatide treatment, whereas they gained weight during insulin glargine treatment; the between-group difference in weight change was statistically significant (LS mean difference, -2.2 [0.3] kg; 95% CI, -2.8 to-1.7; P < 0.001). Both exenatide and insulin glargine were associated with significant reductions from baseline in FSG (-2.9 [0.2] and -4.1 [0.2] mmol/L, respectively; both, P < 0.001), although the reduction was significantly greater with insulin glargine compared with exenatide (LS mean difference, 1.2 [0.3] mmol/L; 95% CI, 0.7 to 1.7; P < 0.001). Compared with insulin glargine, exenatide was associated with significantly lower 2-hour postprandial glucose (PPG) excursions (P < 0.016) and total daily mean glucose excursion (P < 0.001). The proportions of patients reporting nausea during exenatide and insulin glargine treatment were 42.6% and 3.1%, respectively; the proportions reporting vomiting were 9.6% and 3.1%. The incidence of hypoglycemia in the 2 groups was 14.7% and 25.2% (P = NS). CONCLUSIONS: In this open-label, crossover study, treatment with exenatide or insulin glargine for 16 weeks was associated with similar significant improvements from baseline in HbA(1c), independent of treatment order. The improvements in HbA(1c) from baseline did not differ significantly between treatment groups. Exenatide therapy was associated with significant reductions in body weight and PPG excursions compared with insulin glargine, whereas insulin glargine was associated with a significantly greater reduction in FSG compared with exenatide. These findings provide additional information to guide treatment decisions in patients with type 2 diabetes who are potential candidates for either therapy.