We studied the effect of the 5alpha-reductase inhibitor MK-434 on responses to testosterone (T) in orchiectomized (ORX) male Brown Norway (BN) rats aged 13 mo. At 4 wk after ORX or sham surgery, a second surgery was performed to implant pellets delivering 1 mg T/day or placebo pellets. During the second 4 wk of the study, rats received injections of MK-434 (0.75 mg/day) or vehicle injections. Treatment with T elevated serum T to 75% above that for sham animals (P = 0.002) and did not affect serum dihydrotestosterone (DHT) or serum estradiol. T treatment also caused an elevation of prostate T and a marked elevation of prostate DHT. During the second half of the study, ORX rats lost an average of 18.86 +/- 4.62 g body wt. T completely prevented weight loss, and the effect was not inhibited by MK-434 (P < 0.001). ORX produced a nonsignificant trend toward a small (5%) decrease in the mass of the gastrocnemius muscle (P = 0.0819). This trend was also reversed by T, and the effect of T was not blocked by MK-434. T caused a significant 16% decrease in subcutaneous fat that was not blocked by MK-434 (P < 0.05). Finally, T caused a 65% decrease in urine excretion of deoxypyridinoline, a marker of bone resorption, and again the effect was not blocked by MK-434 (P < 0.0001). In contrast, T caused a greater than fivefold increase in prostate mass, and the effect was almost completely blocked by MK-434 (P < 0.0001). This study demonstrates that 5alpha-reductase inhibitors may block the undesirable effects of T on the prostate, without blocking the desirable anabolic effects of T on muscle, bone, and fat.