OBJECTIVES: To determine whether testosterone supplementation improves rehabilitation outcomes in ill older men. DESIGN: A randomized, placebo-controlled, double-blind study. SETTING: A Geriatric Evaluation and Management (GEM) unit based at a university-affiliated Veterans Affairs Medical Center. PARTICIPANTS: Fifteen men aged 65 to 90 years admitted to the GEM for rehabilitation. INTERVENTION: Subjects were randomized to receive weekly intramuscular injections with testosterone enanthate 100 mg or placebo. MEASUREMENTS: Task-specific performance using the Functional Independence Measure (FIM) and grip strength was measured at the onset of the study and at the time of discharge from the GEM. RESULTS: At baseline, FIM scores were similar between the placebo and the testosterone group (73.7 vs 70.7, P =.637), as was grip strength (49.7 vs 55.3 pounds, P =.555). At discharge from the GEM, testosterone-treated patients had improved FLM scores compared with baseline (93.6 vs 70.7; P =.012) and grip strength (68.7 vs 55.3 pounds; P =.033). In the placebo group there was no significant improvement of FIM scores compared with baseline (78.0 versus 73.7; P.686) or of grip strength (48.9 vs 49.7 pounds; P =.686). CONCLUSIONS: Testosterone supplementation may improve rehabilitation outcomes in ill older men.