Multiple endocrine and metabolic consequences of human immunodeficiency virus (HIV) infection exist that may contribute to bone loss in men with the acquired immune deficiency syndrome (AIDS) wasting syndrome. Recent studies suggest that anabolic strategies can increase lean body mass in men with AIDS wasting. Prior studies have not examined the effects of anabolic agents on bone mineral density (BMD) or bone turnover in these men. To determine the effects of testosterone and progressive resistance training on BMD and bone turnover in eugonadal men with AIDS wasting, we randomly assigned 54 eugonadal men with AIDS wasting (weight < 90% IBW or weight loss >10% from preillness baseline) to receive either testosterone enanthate (200 mg/week, im) or placebo and to progressive resistance training (3 times/week) or no training in a 2 x 2 factorial study design for 3 months. The BMD of the lumbar spine, proximal femur, and total body; lean body mass; and fat mass were measured by dual energy x-ray absorptiometry. Total body scans were repeated after 12 weeks of therapy. Baseline bone turnover and BMD were compared with those in 35 age-matched healthy non-HIV-infected control subjects. Compared with controls, lumbar spine BMD (1.021 +/- 0.018 vs. 1.084 +/- 0.025 g/cm(2); P = 0.04) and total hip BMD (0.951 +/- 0.017 vs. 1.070 +/- 0.019 g/cm(2); P < 0.0001) were reduced in men with AIDS wasting. T-scores were lower in men with AIDS wasting at the lumbar spine (-0.62 +/- 0.17 vs. -0.07 +/- 0.23, P = 0.05) and total hip (-0.65 +/- 0.11 vs. +0.20 +/- 0.014, P < 0.0001). Total hip T scores were less than -1.0 in 33% of men with AIDS wasting. Neither the use of protease inhibitors nor the duration of protease inhibitors use correlated with BMD. Serum osteocalcin levels were lower (3.63 +/- 0.29 vs. 4.54 +/- 0.31 nmol/L; P < 0.04) and urinary N-telopeptide excretion was higher (45.4 +/- 4.5 vs. 26.8 +/- 3.0 nmol BCE/mmol creatinine; P = 0.004) in men with AIDS wasting than in controls. Lumbar spine BMD, as assessed on regional total body dual energy x-ray absorptiometry scan, increased over the 12-week treatment period in response to testosterone (+2.4 +/- 1.3 vs. -1.3 +/- 1.0%, testosterone vs. placebo, respectively; P = 0.02), but not in response to training (+0.8 +/- 1.0 vs. +0.4 +/- 1.3%, training vs. no training; P = 0.70). Lumbar spine and total hip BMD are reduced in eugonadal men with AIDS wasting. Biochemical markers of bone turnover suggest that bone formation and bone resorption are uncoupled in these men. Testosterone administration, but not resistance training, over 3 months increases lumbar spine BMD in eugonadal men with AIDS wasting.