Leptin levels in lean adults vary in response to short-term alterations in energy balance. We tested whether leptin responded to short-term changes in energy balance in obese males in a similar manner to lean individuals. We enrolled eight obese, healthy males in a 12-day study composed of four consecutive dietary treatment periods of 3 days each: baseline eucaloric feeding followed by randomized crossover periods of overfeeding (130% of total energy expenditure (TEE)) or underfeeding (70% of TEE), separated by a eucaloric (100% of TEE) washout period. We measured TEE with doubly labeled water prior to baseline. Leptin levels were measured throughout the third day of each treatment and 24-h weighted averaged were calculated. Subjects' ad libitum intake during a breakfast buffet following each treatment period was recorded. During underfeeding, leptin levels decreased by 21 +/- 6% (P < 0.01) from the previous eucaloric period. During overfeeding, leptin levels increased by 25 +/- 11% (P < 0.01) when subjects were underfed first, but did not increase (5 +/- 8%, nonsignificant (n.s.)) when subjects were overfed first. Changes in ad libitum intake from baseline were calculated for each subject after over-, under-, and eucaloric feeding and did not to correlate with the changes in mesor leptin levels from baseline (R(2) = 0.006, n.s). Leptin levels in obese males were acutely responsive to negative energy balance, but not to positive energy balance unless subjects were previously underfed. Consequently, leptin levels in obese males do not respond to changes in energy intake in a manner that would protect against weight gain.