1. Ten female subjects completed two similar experimental procedures (periods 1 and 2) to obtain values of reproducibility of energy intake and 24 h energy expenditure (24hEE) measurements in a whole body indirect calorimeter. The periods consisted of consumption of a provided weight-maintenance diet for 6–8 d, faeces and urine collection during the last 4 d and occupation of the calorimeter during the last 3 d. The daily routine inside the calorimeter simulated a sedentary day in normal life with some physical activity: 8 h sleep, 75 min bicycling and the remaining time spent on sedentary activities. The metabolizable energy (ME) content of the diet (14% energy as protein, 46% energy as carbohydrate, 40% energy as fat) was calculated using food tables. The actual ME intake as well as digestibility and metabolizability of the diet were obtained later by analyses of food, faeces and urine for energy. Three consecutive 24hEE measurements were performed during the stay in the calorimeter in each period. The time interval between the two periods varied from 2 to 24 months. Reproducibility was assessed at group and individual level.
2. Mean digestibility and metabolizability of the diet showed no significant difference between periods. The within-subject coefficient of variation of metabolizability between periods was 1.7%.
3. Mean 24hEE (MJ) over 3 d did not differ between period 1 (8.78 (SD 0.63)) and period 2 (8.73 (SD 0.66)). The within-subject coefficient of variation in mean 24hEE over three successive days between periods was 3.1% but decreased, after deletion of values for subjects who were less adapted to the calorimeter, to 1.9%.
4. The results are discussed with regard to length of trial and the number of subjects required to test a difference in energy metabolism using whole body indirect calorimeters.