Objective: To determine whether the age-related decrease in resting energy expenditure (REE) is explained by variations in body composition. Design: In Study 1, adult subjects (20-70 years) from the Québec Family Study were classified into five different age groups. Body composition was measured by hydrodensitometry to determine fat mass and fat-free mass as predictors of REE. In the youngest group of individuals these predictors were used to plot a reference regression that was then used to predict REE in the other age groups. In Study 2, this issue was investigated in a longitudinal design (6-year follow-up). Subjects were subdivided into three groups and a reference regression was plotted at the beginning of the follow-up and was then used to predict REE 6 years later in the three age groups. Subjects: In Study 1, 627 adults (288 men and 339 women), aged between 20 and 70 years. In Study 2, 191 adults (93 men and 98 women). Results: In Study 1, measured REE was 329, 302, 528 and 636 kJ/day (Po0.0001) below predicted REE at 34, 44, 54 and 64 years, respectively. In Study 2 the most marked deviation from predicted REE in response to the 6-year follow-up in men was observed in young adults (À548 kJ/day, Po0.001) while in women, the largest deviation occurred later in life (À720 kJ/day, Po0.001). Conclusion: Aging is accompanied by a decrease in REE that is significantly greater than what is predicted by variations in body composition. This decrease may reach a mean level of about 500-800 kJ/day.