Four obese and four lean women were studied for 4 or 5 weeks in a metabolic unit to assess their short-term responses to carbohydrate-and protein-containing meals and the effects of beta-adrenergic blockade during both weight maintenance and semistarvation. The study was divided into four periods: a period of weight maintenance; a maintenance diet with propranolol; semistarvation; semistarvation plus propranolol. The low-energy diets contained half the amount of carbohydrate and fat but the protein intake was maintained. The metabolic rate was measured in the fasting state by indirect calorimetry prior to a test meal and then for prolonged periods up to 5.5 h after carbohydrate and 7.5 h after protein test meals. Three lean subjects also spent 23 h in a whole-body calorimeter in each period, when BMR and metabolic rate at different activity levels were measured. Propranolol causes a fall in BMR in both groups on a maintenance diet, but had little further effect on the lower BMR during semistarvation. The protein meal caused a peak increase of 30 percent in oxygen consumption (321 kJ in obese, 257 kJ in lean), two to three times that of the carbohydrate meal (91 kJ) despite the energy intake being 25 percent less in the protein meal. There was no consistent difference in postprandial thermogenesis in obese and lean subjects and the effect was not modified by propranolol or semistarvation. Propranolol caused reduced thermogenesis after a mixed meal in lean subjects.