Circulating concentrations of leptin are closely correlated with body fat mass, and may thus constitute an afferent limb of a system regulating body fatness, with efferent limbs that affect energy expenditure and food intake. We studied 50 subjects (27 males, 23 premenopausal females; 31 never-obese, 19 obese) at usual body weight during active weight loss or weight gain and during the maintenance of body weights 10% above usual (WT + 10%) and 10% and/or 20% below usual body weight (Wt -10% and Wt -20%) to test the hypotheses that the dynamic process of weight change and the maintenance of an altered body weight are associated with significant changes in circulating concentrations of leptin and/or the relationship between fat mass and leptin, and such changes in the plasma concentration of leptin are related to changes in energy expenditure at altered body weight. Subjects were admitted to the Rockefeller University Hospital, and energy metabolism (24-h energy expenditure, resting energy expenditure, thermic effect of feeding, and nonresting energy expenditure) and circulating concentrations of leptin and insulin were examined at various weight plateaus (usual body weight, 10% above usual body weight, 10% below usual body weight, and 20% below usual body weight). Plasma leptin was also measured in some subjects during dynamic periods of weight gain or loss. Though both plasma leptin concentrations and fat mass were significantly correlated with resting energy expenditure, only the correlation of fat mass and energy expenditure remained significant in a multiple stepwise linear regression analysis. Neither absolute nor relative changes in plasma leptin between weight plateaus were significantly correlated with any of the observed changes in energy expenditure. Plasma leptin concentrations were significantly lower during weight loss than during weight maintenance at the same body composition. Plasma leptin concentrations, normalized to fat mass, were significantly lower during the maintenance of a reduced body weight in females and higher during the maintenance of an elevated body weight in males than in the same subjects at usual body weight. At all weight plateaus, plasma leptin concentrations normalized to fat mass were significantly higher in females than in males, but gender was not a significant covariate of the relationship between leptin and energy expenditure. Postabsorptive serum concentrations of insulin was a significant covariate of plasma leptin concentration in males, but not females, at Wt initial and Wt + 10%. Although plasma leptin is significantly reduced during dynamic weight loss compared with static weight maintenance at the same body weight, the lack of correlation between changes in plasma leptin and changes in energy expenditure between weight plateaus suggests that leptin is not the primary signal that mediates the changes of energy expenditure that accompany the maintenance of an altered body weight in humans.