To assess whether long-term habitual exercise affects energy expenditure even on a nontraining day, 24-h energy expenditure (24-h EE) and metabolic rate of sedentary, moderately active, and highly active males (n = 10 per matched group), were measured in a room respirometer on two separate occasions: sedentary and standardized mild-exercise protocols. Twenty-four--hour EE was greatest in the highly active group, second highest in the moderately active group, and lowest in the sedentary group on both experimental days (sedentary day: 9908 +/- 344, 9328 +/- 357, and 8669 +/- 227 kJ/d; exercise day: 11915 +/- 395, 11609 +/- 328, and 11063 +/- 370 kJ/d, respectively). Differences were significant between the 24-h EE (P < 0.01), waking (P < 0.03), and sleeping metabolic rate (P < 0.01) of the highly active group compared with the sedentary group. However, when expressed per unit lean body mass (LBM), group values on both experimental days were not significantly different. Therefore, we found no evidence that habitual exercise, at a high or moderate level, leads to a significant prolonged stimulation of metabolic rate per unit active tissue. However, the increased LBM associated with exercise does increase daily energy expenditure by 8-14%.