The factors responsible for the variability in plasma leptin levels observed among individuals with similar body compositions remain unclear. To examine the impact of dietary variables, we compared the changes in leptin levels induced by fasting and dietary fat restriction with the expected decrease following a significant loss in adipose mass. A 21.4 +/- 3.7% weight loss led to a 76.3 +/- 8.1% decrease in mean plasma leptin level (25.2 +/- 9.3 to 6.1 +/- 3.4 ng/mL, P = 0.0001) in a group of 9 obese males. Despite a weight loss of only 2.6 +/- 0.8%, mean plasma leptin levels fell by 61.9 +/- 25.2% (8.5 +/- 4.5 to 2.4 +/- 0.5 ng/mL, P < 0.01) in 7 nonobese females subjected to 3 days of fasting. Leptin levels in fasted subjects returned to baseline within 12 h of refeeding. Individual high- and low-fat meals given to 19 subjects after an overnight fast had no effect on plasma leptin levels. Reduction in dietary fat content from 37-10% of total calories for 7 weeks was also without effect on plasma leptin levels in these subjects. We conclude that plasma leptin levels primarily reflect total adipose mass, rather than meal consumption or dietary energy source, but that the reduction in leptin levels with ongoing fasting is disproportionate to the reduction in adipose mass. The ability of fasting to deactivate this presumed physiological satiety system may have been advantageous in environments characterized by rapid changes in food availability.