Twenty-four hour whole body indirect calorimetry has been used to study the effects of feeding, during a sedentary test day, isoenergetic diets which varied in fat (3 or 40 per cent of total energy) and carbohydrate (82 or 45 per cent) content. Three groups of women were studied: lean, obese and 'post-obese' after slimming. Energy expenditure was greater in absolute terms in the obese women. Twenty-four hour energy expenditure was lower by only 3-7 per cent when fasting compared to that when fed to achieve energy balance. There were no large differences in energy expenditure between the two diets or between the groups but the thermogenic effect of the high carbohydrate diet was significantly greater than that of the high fat diet (5.8 vs 3.5 per cent of energy expenditure: P less than 0.01). The post-obese tended to have lower energy expenditure per kg FFM than controls when fasting and when high-fat fed, but this pattern was not shown by the obese. Sleeping energy expenditure was particularly low in the post-obese group when high-fat fed. Dirunal variations in RQ appear to show more marked rise in morning RQ from the nocturnal minimum in the obese and post-obese, which might be evidence for an energy-saving mechanism through greater availability of stored dietary carbohydrate.