Excess androgen secretion and exogenous androgen administration may decrease insulin sensitivity and impair glucose tolerance. We examined the responses to an oral glucose tolerance test in 30 normal men before and after 6 weekly injections of androgen administered in a double-blinded study design. The men were randomly assigned to 1 of 4 treatment groups: testosterone enanthate (TE), 100 or 300 mg/week, or 19-nortestosterone decanoate (ND), 100 or 300 mg/week. Serum testosterone levels, measured 2-3 days after the last dose, did not change in the men given 100 mg TE/week, increased 3-fold in those given 300 mg TE/week, and decreased in both ND groups. All four groups had comparable reductions in serum LH levels. Weight increased significantly in all except the 100 mg TE/week group, but there was no change in waist to hip ratio in any group. In spite of the demonstrated biological effects of the doses of steroid administered, androgen administration for 6 weeks did not increase fasting serum glucose or insulin concentrations. There was also no increase in peak serum insulin levels and areas under the insulin and glucose response curves after a 100-g oral glucose load. However, the mean serum insulin area under the curve decreased significantly in the men given 300 mg ND/week. In contrast to the results of studies of 17-alkylated androgens, our results demonstrate that pharmacological doses of testosterone and the administration of 19-nortestosterone for 6 weeks do not impair glucose tolerance or alter insulin secretion in normal men.