In an open-label trial, patients with type 2 diabetes with a suboptimal glycated hemoglobin level while receiving a maximally tolerated dose of metformin and sulfonylurea were randomly assigned to receive biphasic, prandial, or basal insulin. The addition of a single analogue-insulin formulation resulted in a glycated hemoglobin level of 6.5% or less in a minority of patients at 1 year. Regimens of biphasic or prandial insulin had greater efficacy than did the basal regimen but were associated with greater risks of hypoglycemia and weight gain. The addition of a single analogue-insulin formulation resulted in a glycated hemoglobin level of 6.5% or less in a minority of patients at 1 year. Regimens of biphasic or prandial insulin had greater efficacy than did the basal regimen but were associated with greater risks of hypoglycemia and weight gain. Type 2 diabetes mellitus is a progressive condition in which the glycated hemoglobin level rises inexorably over time and the function of beta cells declines. 1 , 2 The maintenance of nearly normal glycemic levels reduces the risk of diabetic complications 3 – 5 but is difficult to achieve, despite the administration of escalating doses of oral antidiabetic drugs, such as metformin, sulfonylureas, and thiazolidinediones. 6 – 8 Most patients eventually require insulin, 6 which usually is added when glycemic control with a regimen of oral antidiabetic agents becomes suboptimal. 9 The addition of insulin can result in a clinically relevant improvement in a patient's glycated hemoglobin level. . . .