To assess the safety and efficacy of the short-acting glucagon-like peptide-1 receptor agonist exenatide on a population of patients with type 2 diabetes (T2D) mostly treated with continuous subcutaneous insulin injection (CSII). A phase 2/3, multicentre, randomized, parallel-group, double-blind, placebo-controlled, 6-month trial was conducted. Patients were randomized to receive subcutaneous (SC) injections of exenatide (10 μg BID) or matched placebo. A total of 46 patients with T2D and elevated HbA1c were randomized (42% of the planned sample size): exenatide (n = 28) and placebo (n = 18). CSII treatment was used by 75% and 89% of patients of the exenatide and placebo groups, respectively. At 6 months, the change in HbA1c was -0.62% ± 0.94% and 0.08% ± 0.81% in the exenatide and placebo groups, respectively (difference, -0.70%; 95% CI [-1.24%; -0.15%], P = .014); body weight and body mass index decreased in the exenatide group (-2.55 ± 3.25 kg and -1.00 ± 1.31 kg/m2 ) and increased in the placebo group (1.29 ± 2.82 kg and 0.46 ± 1.16 kg/m2 ) (observed difference, -3.85 and -1.45, respectively, both P < .001); the postdinner capillary blood glucose value was lower in the exenatide group compared with the placebo group (162.4 ± 80.5 vs. 259.1 ± 94.4 mg/dL, respectively; observed difference, -96.7, P < .01). Hypoglycaemic risk, quality of life and overall safety were not different between the groups, apart from the expected occurrence of digestive effects in the exenatide group. Although we failed to reach our planned sample size, the addition of exenatide treatment 10 μg BID SC in T2D patients with uncontrolled HbA1c despite an intensified insulin regimen, resulted in a significant reduction of HbA1c and body weight with a good overall safety profile and acceptance.