The lower responsiveness to GH in women than in men is probably due to a divergent effect of gonadal steroids. It is unknown, however, how the progressive increase in sex steroid production that occurs during puberty affects this responsiveness. To compare the effects of puberty and sex steroid administration on responsiveness to GH, we used the IGF-I generation test, in which the peak IGF-I level 24 h after a single injection of GH (2 mg/m2) was studied in 117 healthy short subjects (56 females and 61 males). The subjects, aged 8-16 yr, were divided into four groups: prepuberty, early puberty, midpuberty, or pubertal delay. In the latter group, the IGF-I response was determined before and after priming with oral 17beta-estradiol in girls and im testosterone in boys. We also tested for an association between body composition (by dual energy x-ray absorptiometry) and the IGF-I response to GH. The IGF-I increment in response to GH (change in IGF-I from baseline) was correlated with the growth velocity sd score (P < 0.05). Progression throughout puberty was associated with an increase in both baseline IGF-I (P < 0.05) and the IGF-I increment in response to GH (P < 0.05), with no gender difference. Pubertal category (pre-, early, and midpuberty; P < 0.05) and fat percentage (P < 0.05) were the main positive predictors of the IGF-I increment in response to GH, expressed as micrograms per liter as well as sd score, independently of baseline IGF-I. After sex steroid priming, both the GH peak in response to insulin-induced hypoglycemia and baseline IGF-I were increased (P < 0.05, after vs. before sex steroid). However, the IGF-I increment in response to GH decreased after oral 17beta-estradiol (P < 0.05), whereas it was unchanged after testosterone administration. Endogenous gonadal steroid secretion appears to result in increased responsiveness to GH in peripubertal girls and boys. By contrast, exogenous estrogen and testosterone, respectively, produce a relative decrease and no change in responsiveness to GH in similar populations, possibly through the achievement of sex steroid concentrations exceeding physiological ranges for age. Fat percentage was a positive determinant of the responsiveness to GH, suggesting a link between the energy stores and the anabolic action of GH.