A double-blind, placebo-controlled, crossover study on the effects of 4 months' growth hormone (GH) treatment was carried out in 22 GH-deficient adults (8 women, 14 men; mean [SEM] age 23.8 [1.2] years). 1 patient was withdrawn because of oedema. Mean total body weight of the other 21 did not change, whereas mean muscle volume of the thigh, estimated by computerised tomography (CT), was significantly higher after GH than after placebo (70.0 [3.7] vs 66.3 [3.1] ml/0.8 cm cross-sectional slice). The mean adipose tissue volume of the thigh and subscapular skinfold thickness fell significantly during GH treatment. Growth hormone caused a small increase in the isometric strength of the quadriceps muscles and a significant rise in exercise capacity (60.8 [7.2] vs 54.2 [6.6] kJ). The heart rate both at rest and after maximum exercise was low during the placebo period and increased significantly during GH treatment. Blood pressure and echocardiographic wall mass of the left ventricle did not change during the study. Growth hormone increased both mean glomerular filtration rate and renal plasma flow from a subnormal level on placebo to a level comparable with that of an age-matched control group. The filtration fraction did not change. Urinary albumin excretion was in the low normal range and was not affected by GH treatment. Finally, GH treatment normalised mean circulating levels of insulin-like growth factor 1 (IGF-1), which were low after the placebo period (96  micrograms/l placebo; 224  micrograms/l GH). These findings suggest that GH, in a conventional replacement dose, has several potentially beneficial effects in GH-deficient adults and therefore encourage future long-term trials.