Megestrol acetate (MA) (Bristol-Myers Squibb Co., Princeton, NJ) increases weight gain in AIDS and cancer patients and in age-related cachexia; however, the weight gain is predominately fat. We determined if adding resistance exercise and/or testosterone (T) replacement to MA administration would result in a more favorable body composition change than MA alone. Thirty older men (aged 67.0 +/- 5.8) completed this 12-wk study. All subjects received MA and were randomly assigned to one of the following groups: placebo (P) injections, resistance training (RT) and P (RT + P), weekly injections of T (100 mg/wk) or, RT and T (RT + T). The mean increase in body weight for all groups combined was 3.8 kg (P < 0.0001), but this increase was not different between groups. There was a significant interaction for the change in thigh muscle cross-sectional area (P = 0.0006). Thigh muscle cross-sectional area was significantly reduced from baseline by 5.20 [1.62] cm(2) (P = 0.05) in P which was not prevented in T [-4.44 (1.66) cm(2) from baseline; P = 0.04]. RT prevented this decline [+0.61 (1.41) cm(2) from baseline]. Muscle cross-sectional area increased 4.51 (1.69) cm(2) from baseline in RT + T (P = 0.002 vs. P and P = 0.002 vs. T). Despite significant weight gain, MA appears to have an antianabolic effect on muscle size even when combined with T replacement. Resistance exercise attenuated this reduction in muscle mass and when combined with T had an anabolic effect on muscle mass.