Several trials with growth hormone (GH) replacement therapy in adults with GH deficiency have been conducted during the last 10 years. Beneficial effects of treatment on bone density, physical capacity, body composition, lipid profile and quality of life have been reported. It has long been known that GH secretion is greater in women than in men, despite similar reference ranges of serum insulin-like growth factor (IGF)-I in adult men and women. It has also been reported that sex steroids influence not only GH secretion but also the local synthesis of IGF-I in target tissues and the expression of the GH receptor in various other tissues. However, it has been acknowledged only recently that there is a clinically significant gender difference in the response to GH treatment in adults with GH deficiency and, consequently, a need to adjust the dose of recombinant human GH (rhGH). We report the results of a placebo-controlled study in 36 men and women with GH deficiency who received the same dose of rhGH per body surface area (1.25 U/m2 per day) for 9 months. We observed significantly greater responses in male patients than in female patients with regard to the changes in serum levels of IGF-I, body composition and biochemical markers of bone metabolism. When these patients continued to receive GH replacement therapy for an additional 24 months, the dose of rhGH was adjusted to the serum levels of IGF-I. As a result, the dose administered to the male patients was reduced to nearly half that given to the female patients (1.0 vs 1.9 U/day) and the serum levels of IGF-I and of biomarkers of bone turnover increased to the same extent in patients of both sexes. However, an increase in bone density of the hip and the lumbar spine after a total of 33 months of rhGH treatment was observed only in the male patients; no significant changes in bone density were found in the female patients. The reason for the observed difference in GH response between men and women with GH deficiency is not known, although the different sex steroid pattern cannot be excluded as a contributing factor.