The thermic effect of glucose was investigated in nine obese and six lean subjects in whom the same rate of glucose uptake was imposed. Continuous indirect calorimetry was performed for 240 min on the supine subject. After 45 min, 20% glucose was infused (609 mg/min) for 195 min and normoglycemia was maintained by adjusting the insulin infusion rate. At 2 h, propranolol was infused (bolus 100 micrograms/kg; 1 microgram/kg X min) for the remaining 75 min. To maintain the same glucose uptake (0.624 g/min), it was necessary to infuse insulin at 3.0 +/- 0.6 (leans) and 6.6 +/- 1.2 mU/kg X min (obese) (P less than 0.02). At this time, glucose oxidation was 0.248 +/- 0.019 (leans) and 0.253 +/- 0.022 g/min (obese) (NS), and nonoxidative glucose disposal was 0.375 +/- 0.011 and 0.372 +/- 0.029 g/min, respectively. Resting metabolic rate (RMR) rose significantly by 0.13 +/- 0.02 kcal/min in both groups, resulting in similar thermic effects, i.e., 5.5 +/- 0.7% (leans) 5.4 +/- 0.9% (obese) (NS) and energy costs of glucose storage 0.35 +/- 0.06 and 0.39 +/- 0.09 kcal/g (NS), respectively. With propranolol, glucose uptake and storage remained the same, while RMR fell significantly in both groups, with corresponding decreases (P less than 0.05) in the thermic effects of glucose to 3.7 +/- 0.6% and 2.9 +/- 0.8% (NS) and the energy costs of glucose storage 0.23 +/- 0.04 and 0.17 +/- 0.05 kcal/g (NS) in the lean and obese subjects, respectively. These results suggest that the defect in the thermic effect of glucose observed in obese subjects is due to their insulin resistance, which is responsible for a lower rate of glucose uptake and hence decreased rate of glucose storage, which is an energy-requiring process.