Two experiments have been performed to establish whether exercise-training has an influence on resting metabolic rate (RMR). In a first study, RMR was measured in a cohort of 59 individuals comprising 20 trained and 39 non-trained subjects. The absolute level of RMR in trained subjects exceeded by 11 percent that observed in the non-trained individuals (P less than 0.01). When comparing regression lines of RMR versus FFM between the two groups, the intercept with the Y axis (RMR values) was also significantly higher in trained subjects (P less than 0.01). The second experiment was conducted to find out whether the trend for an elevated RMR noted in athletes could be reproduced in obese persons engaging in an exercise-training program. Eight moderately obese women were submitted to an 11-week training programme, including 5 hours of aerobic exercise per week performed at a mean intensity of about 50 percent VO2 max. The results showed that exercise-training induced a significant rise in RMR which corresponded to 8 percent of pretraining value in kcal/kg FFM/min (P less than 0.01). Thus, data reported here suggest that aerobic exercise-training is associated with an elevated RMR per unit of fat free mass in both lean and moderately obese individuals.