In subjects who maintain a constant body mass, the increased energy expenditure induced by exercise must be compensated by a similar increase in energy intake. Since leptin has been shown to decrease food intake in animals, it can be expected that physical exercise would increase energy intake by lowering plasma leptin concentrations. This effect may be secondary either to exercise-induced negative energy balance or to other effects of exercise. To delineate the effects of moderate physical activity on plasma leptin concentrations, 11 healthy lean subjects (4 men, 7 women) were studied on three occasions over 3 days; in study 1 they consumed an isoenergetic diet (1.3 times resting energy expenditure) over 3 days with no physical activity; in study 2 the subjects received the same diet as in study 1, but they exercised twice daily during the 3 days (cycling at 60 W for 30 min); in study 3 the subjects exercised twice daily during the 3 days, and their energy intake was increased by 18% to cover the extra energy expenditure induced by the physical activity. Fasting plasma leptin concentration (measured on the morning of day 4) was unaltered by exercise [8.64 (SEM 2.22) 7.17 (SEM 1.66), 7.33 (SEM 1.72) 1 microg x l(-1) in studies 1, 2 and 3, respectively]. It was concluded that a moderate physical activity performed over a 3-day period does not alter plasma leptin concentrations, even when energy balance is slightly negative. This argues against a direct effect of physical exercise on plasma leptin concentrations, when body composition is unaltered.