Short-term exercise (<60 min) studies suggest that leptin concentrations are not acutely affected in healthy males and females. Most reports of reductions in serum leptin may be attributed to circadian rhythms or hemoconcentration. For long-term (> or =60 min) exercise, a reduction in leptin concentrations reported from 1 to 3 hr of running or cycling has been attributed to diurnal reduction in circulating leptin, independent of exercise. Exercise that produces a sufficient energy imbalance (kilocalorie intake versus kilocalorie expenditure) suppresses 24-hr mean and amplitude of the diurnal rhythm of leptin in women. Suppression of leptin concentrations may be counterbalanced by feeding and may explain consistent reports of reductions in leptin concentrations following extreme bouts of exercise such as marathons or ultramarathons. In addition, leptin concentrations are reduced 48 hr after long-term aerobic exercise and long-term resistance exercise is associated with delayed leptin reduction 9 hr postexercise. Training studies have documented that short-term exercise training (< or =12 weeks) does not affect leptin levels, with the exception of patients with type 2 diabetes. Exercise training protocols that result in reduced fat mass will lower leptin concentrations, thus, most investigators have reported leptin concentrations after accounting for fat loss. There are disparate findings concerning long-term (>12 weeks) training studies, with a number of studies finding no effect of training on leptin concentrations other than effects induced by fat loss, and other studies finding reductions in leptin concentrations after accounting for fat loss. Exercise training-induced reductions in leptin levels have been attributed to alterations in energy balance, improvements in insulin sensitivity, alterations in lipid metabolism, and unknown factors. Hormone replacement does not seem to affect leptin adaptations to training. Patients with type 2 diabetes show delayed effects of short-term resistance exercise on leptin concentrations, reduced leptin levels with long-term training, and appear to be more sensitive to training-induced leptin adaptations than other populations.