Serum leptin concentrations increase during childhood in both sexes. During sexual maturation, levels rise further in girls, but decrease in boys. These data suggest that testosterone either directly suppresses leptin levels or induces changes in body composition that result in lower leptin concentrations. To examine further the relationship between sex steroids and leptin, we performed a longitudinal study in children with central precocious puberty (28 girls and 12 boys) before, during, and after discontinuation of GnRH agonist-induced pituitary-gonadal suppression. Nighttime and daytime leptin levels were measured to determine whether the activity of the pituitary-gonadal axis affects their diurnal variation. In the boys, suppression of testosterone increased leptin levels, whereas resumption of puberty was associated with decreased leptin levels [3.5 +/- 0.8 vs. 9.5 +/- 3.1 ng/dL (P = 0.005) and 12.2 +/- 4.5 vs. 7.0 +/- 2.6 ng/dL (P = 0.012), respectively]. Serum leptin levels did not change in the girls with alteration of the pituitary-ovarian axis and consistently exceeded those in boys. Nighttime levels were consistently greater than daytime values by an average of 38.3% in the girls and 29.4% in the boys. These serial observations during reversible pituitary-gonadal suppression suggest that testosterone decreases leptin concentrations, but that estrogen, at least in this childhood model, has no discernible effect. In addition, our data indicate that the presence of the diurnal rhythm in leptin concentrations is independent of the state of the reproductive axis.