A reduced metabolic rate in the etiology of obesity has been a subject of controversy. The prediction of the energy requirements for the obese using reference values may therefore be distorted. In order to examine this possibility, resting metabolic rate (RMR) while the subject was sitting comfortably in a chair was measured in a total of 134 moderately obese and normal-weight subjects (68 women aged 20 to 71 with a mean of 53.1 and 66 men aged 20 to 63 with a mean of 36.5). RMR per kg of body weight was significantly lower in the female obese subjects, but not in the male obese subjects. There was no evidence of difference in RMR between obese and normal-weight subjects in either sex when RMR was indexed with fat-free mass (FFM), indicating no substantial decrease in the metabolism due to obesity. Multiple regression analyses indicate that standardization of RMR by FFM eliminates the apparent difference in RMR between the sexes, and the diminution of RMR with age was not observed. While the best and logical prediction of RMR is to use FFM, regression analyses suggest an alternative way of predicting RMR by an incorporation of subscapular skinfold thickness to adjust the different body composition in lean and obese subjects. Prediction equations of postprandial RMR (kcal/24-h) while sitting are RMR = 24.5 x FFM(kg) + 303.7, and RMR = 22.7 x weight(kg) - 13.6 x SSF (subscapular skinfold: mm) + 350.6. Problems in predicting RMR are discussed.