We administered testosterone transdermally to six hypogonadal men by applying a thin flexible polymeric membrane containing testosterone to the scrotum. Each man wore a placebo membrane and three doses (5, 10, and 15 mg) of testosterone-containing membranes for 22 h/day. Each dose was worn for 1 week, and one dose was worn a second week. Blood was sampled frequently for one 22 h period on the seventh day of each treatment period. After the application of a membrane, the serum testosterone concentration rose rapidly, reached a peak in 2-3 h, and decreased slowly to 60-80% of the peak value by 22 h. The mean (+/- SE) 22-h average testosterone concentration during the wearing period was dependent on the testosterone content of the membrane (placebo, 135 +/- 38 ng/dl; 5 mg, 348 +/- 66 ng/dl; 10 mg, 455 +/- 77 ng/dl; and 15 mg, 624 +/- 65 ng/dl; P less than 0.001, by analysis of variance). When the same dose was worn twice, the mean coefficient of variation was 13.9%. We conclude that the transdermal application of testosterone to hypogonadal men reproducibly raises their serum testosterone concentrations to within the normal range, and that it, therefore, warrants evaluation as a treatment for male hypogonadism.