The applicability of body composition as estimated by the bioimpedance method to predict energy expenditure (EE) was studied. Ten healthy subjects underwent measurement of body composition and 24-h energy expenditure (24-h EE) twice in a respiration chamber on a fixed program. The 24-h EE and its components, sleeping EE (SEE), basal EE (BEE), and daytime EE, for an individual were very reproducible (coefficient of variation 2.3%, 1.4%, 5.0%, & 3.1%, respectively). The variability of 24-h EE among subjects was 11.4% but only 4.1% when adjusted for differences in lean body mass (LBM). LBM was the best determinant of 24-h EE, BEE, and SEE and accounted for 91-93% of the interindividual variance of EE. The prediction equations were 24EE (kcal/d) = 390 + 33.3 LBM (r2 = 0.93, P = 0.000001), SEE (kcal/h) = 9.8 + 1.1 LBM (r2 = 0.92, P = 0.000001), and BEE (kcal/h) = -3.1 + 1.35 LBM (r2 = 0.91, P = 0.000002). In conclusion, 24EE, BEE, and SEE can be predicted with a high degree of precision from LBM as estimated by bioimpedance in normal-weight subjects.