An understanding of energy expenditure in hospitalized patients is necessary to determine optimal energy supply in the care of individuals who require nutritional support. A review was conducted of 19 studies in which resting energy expenditure (REE) had been measured using indirect calorimetry and compared with estimated basal energy expenditure (BEE) from the Harris-Benedict equation. Studies of patients with burns, head injuries, and fever were excluded because REE is known to be increased in these conditions. The studies reported data on 1256 patients with the following diagnoses: postoperative (28%), trauma or sepsis (26%), cancer (18%), pulmonary disease (9%), cardiovascular disease (2%), miscellaneous (9%), and unspecified (6%). The average REE in the 19 studies was 113% of the BEE. The mean +/- SD REE/BEE ratio was higher in 11 studies in which the REE was measured during feeding than in 5 studies in which the measurement was made during fasting (117% +/- 3% vs 105% +/- 4%; P = .047). In those 11 studies, overfeeding may have contributed to higher REE values than otherwise would have been observed. Some evidence indicated that the REE/BEE ratio is higher in more severe illness, but results were inconsistent. Unfortunately, little information is available concerning total energy expenditure, which includes the contribution of physical activity. It appears that most patients can be fed adequately with energy equal to 100% to 120% of estimated BEE. Hypoenergetic feeding may be appropriate in some overweight and obese individuals. Additional research in hospitalized patients on total energy expenditure and on the relationship between severity of illness and energy expenditure is needed.