We compared three independent techniques for measurement of total energy expenditure (TEE) in human subjects: 1) weight-maintaining energy intake (dietTEE), 2) 24-h chamber calorimetry (chamberTEE), and 3) differential elimination rates 2H2O and H(2)18O (isotopeTEE). Twenty-three healthy adult in-patients [19 never obese (NO), 2 obese (OB), and 2 formerly-obese (RO); 9 female, 14 malel] ingested a liquid formula diet (40% of calories as fat, 45% carbohydrate, 15% protein), the volume of which was adjusted until body weight was stable for at least 14 days. Body composition was then determined by hydrodensitometry, isotope dilution, and dual photon beam absorptiometry (DXA). The thermic effect of feeding (TEF) and resting energy expenditure [REE; measured before arising (dietREE) and after arising (chamberREE)] were determined by indirect calorimetry. Non-resting energy expenditure (NREE) was calculated as NREE = TEE - (REE + TEF). Subjects then gained or lost 10% of their body weight and were restudied as described above. All measures of TEE were significantly correlated (dietTEE vs. chamberTEE r2 = 0.75; dietTEE vs. isotopeTEE r2 = 0.88; isotopeTEE vs. chamberTEE r2 = 0.73; P < 0.0001). ChamberTEE (mean +/- SE = 2,107 +/- 64 kcal/day) was approximately 20% lower than either dietTEE (2,536 +/- 94 kcal/day, P < 0.0001) or isotopeTEE (2,564 +/- 83 kcal/day, P < 0.0001). When data were normalized to metabolic mass, weight gain of 10% was associated with significant increases in dietTEE (P < 0.005) and isotopeTEE (P < 0.05) but not chamberTEE; weight loss of 10% was associated with significant reductions in dietTEE (P < 0.005) and isotopeTEE (P < 0.05) but not chamberTEE. We conclude that measures of energy expenditure obtained in a highly controlled environment by caloric titration (dietTEE) or differential excretion rates of 2H2O and H(2)18O (isotopeTEE) are not significantly different and that measurements of TEE obtained in a respiratory chamber (chamberTEE) are significantly lower than dietTEE or isotopeTEE, probably largely due to limitations on physical activity in the chamber.