Circulating concentrations of leptin ([leptin]) vary directly with body mass index and percentage body fat, and may thus constitute an afferent limb of a system regulating body fatness. We tested the hypotheses that: 1) Plasma [leptin] vary more directly with absolute fat mass than with fractional body fatness per se: and 2). The relationship between fat mass and [leptin] is significantly affected by gender and by menopausal status. [Leptin] in the post-absorptive state was examined in 67 subjects (26 male, 20 premenopausal female, 21 postmenopausal females; 43 never-obese, 24 obese) at usual body weight. Body composition was determined by hydrodensitometry, and [leptin] was determined by a double antibody ELISA assay. In male and pre-menopausal female subjects, subcutaneous adipose tissue aspirations were performed for determination of adipocyte volume by the osmium fixation method, and a 3 hour oral glucose tolerance tests was performed. At usual body weight, ([leptin]) was better correlated with absolute fat mass than with body mass index (BMI) or percentage body fat. BMI and % body fat did not account for any of the variance in [leptin] beyond that attributable to FM, per se. The regression equations relating FM to [leptin] did not differ significantly between obese and never-obese subjects. [Leptin] and fasting serum insulin concentrations were significantly correlated in males only. [Leptin] was significantly higher in pre- and post-menopausal females compared to males, even when [leptin] was corrected for differences in body composition (pre-menopausal females > post-menopausal females > males). While plasma [leptin], corrected for FM, declines significantly in women post-menopause, this decline is not sufficient to account for the striking sexual dimorphism in the relationship of leptin to fat mass. This sexual dimorphism is apparently also due, in part, to a suppressive effect of circulating androgens on [leptin].