To review the major data collected over the past 8 yr regarding energy expenditure in relationship to obesity and the development of obesity in the Pima Indian population. The different components of 24-h energy expenditure (i.e., the RMR, the thermic effect of food, and the energy cost of spontaneous physical activity) were measured in a respiratory chamber after a few days on a weight-maintenance diet. Fat-free body mass, the major determinant of RMR, explains 82% of its variance. However, an extra 7% of the variance in RMR observed between people is related to family membership. The variability of RMR for a given body size and composition is of importance, because a low metabolic rate is a major risk factor for weight gain in man. Also, genetic factors seem to be the cause of the familial aggregation of metabolic rate in man. The high prevalence of obesity and NIDDM in the Pima Indian community might be the consequence of a "thrifty genotype." The increasing evidence that obesity cannot always be attributed to gluttony and sloth forces us to consider obesity as a "real metabolic disease" that needs to be treated as such, using new behavioral and pharmacological therapies.