For the past few years, interest has focused on the combination of insulin and sulfonylurea in the search for effective treatments for non-insulin-dependent diabetes mellitus (NIDDM) patients who fail on oral hypoglycemic agents. Although several studies have demonstrated beneficial effects of such therapy in NIDDM patients, the average effect is small and is observed in only about half of the patients. However, there are several problems with most previous studies, including small sample size, selection of patients, and simultaneous use of several end points. First, residual beta-cell function has been considered to be a prerequisite for a beneficial effect of combination therapy. Therefore, most studies have failed to demonstrate improved glycemic control after adding sulfonylurea to insulin therapy in patients with insulin-dependent diabetes mellitus. Inclusion of patients with impaired beta-cell function will therefore attenuate the effect of combination therapy. Second, most studies have used glycemic control as an end point. Nevertheless, the insulin dose has been reduced by 20-30% to avoid hypoglycemia after adding sulfonylurea to insulin. Thus, the comparison has been made between treatments with a smaller insulin dose with sulfonylurea and a larger insulin dose without sulfonylurea. The patient most likely to benefit from combination therapy is slightly obese, has had NIDDM for a relatively short period, and has preserved beta-cell function. In such a patient, combined insulin-sulfonylurea therapy predominantly stimulates basal insulin secretion, resulting in more effective suppression of hepatic glucose production and lower fasting plasma glucose. The side effects are few, most notably more frequent but mild hypoglycemic reactions.