Debate exists over the impact of caloric restriction (CR) on the level of energy expenditure. At the whole animal level, CR decreases metabolic rates but in parallel body mass also declines. The question arises whether the reduction in metabolism is greater, smaller or not different from the expectation based on body mass change alone. Answers to this question depend on how metabolic rate is normalized and it has recently been suggested that this issue can only be resolved through detailed morphological investigation. Added to this issue is the problem of how appropriate the resting energy expenditure is to characterize metabolic events relating to aging phenomena. We measured the daily energy demands of young and old rats under ad libitum (AD) food intake or 40% CR, using the doubly labeled water (DLW) method and made detailed morphological examination of individuals, including 21 different body components. Whole body energy demands of CR rats were lower than AD rats, but the extent of this difference was much less than expected from the degree of caloric restriction, consistent with other studies using the DLW method on CR animals. Using multiple regression and multivariate data reduction methods we built two empirical predictive models of the association between daily energy demands and body composition using the ad lib animals. We then predicted the expected energy expenditures of the CR animals based on their altered morphology and compared these predictions to the observed daily energy demands. Independent of how we constructed the prediction, young and old rats under CR expended 30 and 50% more energy, respectively, than the prediction from their altered body composition. This effect is consistent with recent intra-specific observations of positive associations between energy metabolism and lifespan and theoretical ideas about mechanisms underpinning the relationship between oxygen consumption and reactive oxygen species production in mitochondria.