BACKGROUND: Patients with type 2 diabetes are often obese and require large doses of insulin to achieve glycemic control. Weight gain often accompanies insulin therapy and results in increasing insulin requirements. OBJECTIVE: To evaluate the efficacy of metformin in combination with insulin in patients with type 2 diabetes poorly controlled with insulin therapy alone. DESIGN: Randomized, double-blind, placebo-controlled trial. SETTING: Outpatient diabetes clinic at a university medical center. PATIENTS: 43 patients with poorly controlled type 2 diabetes who were receiving insulin therapy. INTERVENTION: Patients were randomly assigned to receive placebo or metformin in combination with insulin for 24 weeks. RESULTS: Hemoglobin A1c levels decreased by 2.5 percentage points (95% CI, 1.8 to 3.1 percentage points) in the metformin group, a significantly greater change (P = 0.04) than the decrease of 1.6 percentage points in the placebo group. Average final hemoglobin A1c levels were 6.5% in the metformin group and 7.6% in the placebo group (difference, 11%). For patients who received placebo, the insulin dose increased 22.8 units (CI, 11 to 44 units) or 29% more than did the dose for patients who received metformin (P = 0.002); for these patients, the insulin dose decreased slightly. Patients in the placebo group gained an average of 3.2 kg of body weight (CI, 1.2 to 5.1 kg); patients in the metformin group gained an average of 0.5 kg of body weight (P = 0.07). Total cholesterol and low-density lipoprotein cholesterol levels decreased in both groups. High-density lipoprotein cholesterol and triglyceride levels did not change. CONCLUSIONS: The addition of metformin to insulin therapy resulted in hemoglobin A1c concentrations that were 10% lower than those achieved by insulin therapy alone. This improvement in glycemic control occurred with the use of 29% less insulin and without significant weight gain. Metformin is an effective adjunct to insulin therapy in patients with type 2 diabetes.