Thirty-three diabetic men were instructed in the use of phentolamine-papaverine injections for the treatment of erectile dysfunction over a two-year period. Of these, 12 reported a satisfactory response and 21 reported a nonsatisfactory response. The responders and the nonresponders were retrospectively studied to identify characteristics that would predict a satisfactory response. No difference was found between the two groups in the duration of diabetes, the presence of retinopathy, neuropathy, nephropathy, peripheral vascular disease, or ischemic heart disease. The utilization of insulin, the prevalence of type II diabetes, or the use of drugs which would cause impotence, did not differ between the two groups. There was no difference in the serum testosterone levels between the two groups. Age was the only predictive factor. Only 1 of 14 patients over age sixty had a satisfactory response to treatment while 11 of 19 patients under age sixty had a satisfactory response. Five of the responders and 2 of the nonresponders proceeded to penile implant surgery and reported satisfactory results. While older diabetic men may choose a trial of intracorporeal injections, they should be counselled regarding the high failure rate and alternative forms of therapy.