OBJECTIVE: We sought to determine whether, and if so the mechanism by which, testosterone replacement improves carbohydrate tolerance. RESEARCH DESIGN AND METHODS: Fifty-five elderly men with relative testosterone deficiency ingested a labeled mixed meal and underwent a frequently sampled labeled intravenous glucose tolerance test before and after either placebo or treatment with testosterone patch (5 mg/day) for 2 years. RESULTS: Despite restoring bioavailable testosterone to values observed in young men, the change (24 months minus baseline values) in fasting and postprandial glucose, insulin, and C-peptide concentrations and meal appearance, glucose disposal, and endogenous glucose production were virtually identical to those observed after 2 years of placebo. The change over time in insulin and C-peptide concentrations post-intravenous glucose injection also did not differ. Furthermore, the change over time in insulin action and glucose effectiveness (measured with the unlabeled and labeled "oral" and "intravenous" minimal models), as well as insulin secretion and hepatic insulin clearance (measured with the C-peptide model), did not differ in the testosterone and placebo groups. CONCLUSIONS: We conclude that 2 years of treatment with testosterone in elderly men does not improve carbohydrate tolerance or alter insulin secretion, insulin action, glucose effectiveness, hepatic insulin clearance, or the pattern of postprandial glucose metabolism. Thus, testosterone deficiency is unlikely the cause of the age-associated deterioration in glucose tolerance commonly observed in elderly men.