Studies in animals and humans using supraphysiological doses of dehydroepiandrosterone (DHEA) reported significant changes in body composition and carbohydrate metabolism. To investigate the metabolic action of a physiological DHEA replacement dose, we studied 24 women with adrenal insufficiency (AI; mean +/- SD age, 42.3 +/- 9.3 yr; duration of disease, 9.2 +/- 8.4 yr; body mass index, 23.4 +/- 4.0 kg/m(2)) in a double blind, placebo-controlled, randomized, cross-over design. They received 50 mg DHEA/day and placebo orally for 4 months each, with a 1 -month washout period. Measurements included fasting serum glucose, insulin, leptin, bone markers, anthropometric parameters determined by bioimpedance analysis, and exercise capacity as assessed by an incremental cycling test. DHEA did not induce any change in body mass index (placebo vs. DHEA, 23.3 +/- 4.1 vs. 23.2 +/- 3.9 kg/m(2); P = 0.39), parameters of body composition, or exercise capacity. However, compared with placebo, DHEA replacement led to a significant decrease in serum leptin (absolute change after 4 months, DHEA vs. placebo, -5.3 +/- 8.0 vs. 1.1 +/- 5.7 ng/mL; P = 0.01). This is most likely the result of the DHEA-induced normalization of circulating androgens. DHEA had no effect on fasting glucose, insulin, or the glucose/insulin ratio. Compared with placebo, serum osteocalcin increased slightly, but significantly, during DHEA treatment (absolute change after 4 months DHEA vs. placebo, +1.6 +/- 5.3 vs. -1.2 +/- 6.2 ng/mL; P = 0.02). However, urinary cross-links excretion did not change. In conclusion, replacement of DHEA in a physiological dose in patients with pathological DHEA deficiency does not have a significant effect on carbohydrate metabolism, body composition, or exercise capacity. The biological relevance of the changes in leptin and osteocalcin levels remains to be determined.