OBJECTIVE To compare the relative efficacy, risks, and benefits of insulin with glyburide in achieving normoglycemia in non-insulin-dependent diabetes mellitus. METHODS Randomized, double-blind, placebo-controlled trial with a 9-month treatment period. METHODS University hospital. METHODS Thirty-one patients with non-insulin-dependent diabetes mellitus who did not have normal glucose control with diet alone. METHODS Once-per-day NPH insulin and placebo glyburide, or glyburide and once-per-day placebo insulin injection. Active drug and placebo adjusted in parallel to achieve fasting plasma glucose level less than 6.4 mmol/L (115 mg/dL) without hypoglycemia. RESULTS Insulin and glyburide produced similar improvement in fasting blood glucose levels and hemoglobin A1c concentrations, similar frequencies of mild symptomatic hypoglycemia, and similar weight gain despite dietary reinforcement. Triglyceride and cholesterol levels decreased and high-density lipoprotein cholesterol and ratios of high-density lipoprotein to total cholesterol increased in both groups, with a significantly greater improvement in high-density lipoprotein cholesterol and ratio of high-density lipoprotein total cholesterol in patients treated with insulin. CONCLUSIONS Therapy with glyburide or once-per-day NPH insulin provides a similar degree of glucose control in patients with non-insulin-dependent diabetes mellitus. Insulin may have a relative advantage in that it is associated with higher levels of high-density lipoprotein cholesterol and a higher ratio of high-density lipoprotein to total cholesterol.