We performed continuous indirect calorimetry for 24 h on nine occasions in small premature infants. Oxygen consumption, carbon dioxide production, respiratory quotient, and energy expenditure were calculated for each 2-h period. The mean energy expenditure during the first 6 h was within 6.5% of the mean for the whole 24-h period in all but one case. The mean error in estimating total daily energy expenditure from 6-h measurements was 0.9%. Because positive and negative errors tend to offset each other, we also calculated the mean absolute error, which was 5.6%. The mean coefficient of variation in energy expenditure among the 2-h periods was 11.0%. The mean coefficients of variation in oxygen consumption, carbon dioxide production, and respiratory quotient were 12.8, 9.9, and 14.1%, respectively. Total daily energy expenditure of small premature infants can be estimated from measurements as short as 6 h with sufficient accuracy for most purposes.