We analyzed the effect of growth hormone replacement therapy (36 months) analyzed at a dose adjusted to maintain serum insulin-like growth factor-I level between the median and the upper end of the age-related reference range on bone mineral density, body composition, and carbohydrate metabolism with respect to gender and age in 20 adult patients (9 women, 11 men, mean age: 43 years, range: 21-61 years). The lumbar and femoral T-score was increased after 12 and after 18 months of therapy respectively in men (p < 0.001 and p = 0.002), but did not changed significantly in women. The increase of femoral T-score was greater in young men (< or = 45 years, n = 6) than old men (> 45 years, n = 5, p < 0.001). Body fat was lower in men than in women after 6 months (p = 0.002). The waist/hip ratio only decreased in women (p = 0.044). The waist circumference decreased in both genders after 6 months of therapy (p < 0.001), but more markedly in females than in males (p < 0.05). The sum of skinfold thicknesses was reduced in males after 6 months of therapy (p < 0.001). Changes in body composition parameters measured were independent of age. The glycosylated hemoglobin increased without sex or age difference after 12 months of initiation of therapy (p < 0.001), but fasting glucose and insulin levels did not change during the therapy. Our results indicate that the effect of growth hormone replacement on bone mineral content in adults is age- and gender-dependent, gender dependent on body composition, but independent of age and gender on carbohydrate metabolism.