Although testosterone levels and muscle mass decline with age, many older men have serum testosterone level in the normal range, leading to speculation about whether older men are less sensitive to testosterone. We determined the responsiveness of androgen-dependent outcomes to graded testosterone doses in older men and compared it to that in young men. The participants in this randomized, double-blind trial were 60 ambulatory, healthy, older men, 60-75 yr of age, who had normal serum testosterone levels. Their responses to graded doses of testosterone were compared with previous data in 61 men, 19-35 yr old. The participants received a long-acting GnRH agonist to suppress endogenous testosterone production and 25, 50, 125, 300, or 600 mg testosterone enanthate weekly for 20 wk. Fat-free mass, fat mass, muscle strength, sexual function, mood, visuospatial cognition, hormone levels, and safety measures were evaluated before, during, and after treatment. Of 60 older men who were randomized, 52 completed the study. After adjusting for testosterone dose, changes in serum total testosterone (change, -6.8, -1.9, +16.1, +49.5, and +101.9 nmol/liter at 25, 50, 125, 300, and 600 mg/wk, respectively) and hemoglobin (change, -3.6, +9.9, +20.9, +12.6, and +29.4 g/liter at 25, 50, 125, 300, and 600 mg/wk, respectively) levels were dose-related in older men and significantly greater in older men than young men (each P < 0.0001). The changes in FFM (-0.3, +1.7, +4.2, +5.6, and +7.3 kg, respectively, in five ascending dose groups) and muscle strength in older men were correlated with testosterone dose and concentrations and were not significantly different in young and older men. Changes in fat mass correlated inversely with testosterone dose (r = -0.54; P < 0.001) and were significantly different in young vs. older men (P < 0.0001); young men receiving 25- and 50-mg doses gained more fat mass than older men (P < 0.0001). Mood and visuospatial cognition did not change significantly in either group. Frequency of hematocrit greater than 54%, leg edema, and prostate events were numerically higher in older men than in young men. Older men are as responsive as young men to testosterone's anabolic effects; however, older men have lower testosterone clearance rates, higher increments in hemoglobin, and a higher frequency of adverse effects. Although substantial gains in muscle mass and strength can be realized in older men with supraphysiological testosterone doses, these high doses are associated with a high frequency of adverse effects. The best trade-off was achieved with a testosterone dose (125 mg) that was associated with high normal testosterone levels, low frequency of adverse events, and significant gains in fat-free mass and muscle strength.