The objective of this study was to evaluate resting energy expenditure (REE) in spontaneously breathing and artificially ventilated burn patients during the entire intensive care period. In 27 patients with 51 +/- 20% body surface area burned (BSAB) the REE was determined via indirect calorimetry. Three groups were formed according to the mortality prognosis index of Zawacki et al. In groups A, B, and C the predicted mortality rates were <20%, 20% to 80%, and >80%, respectively. The frequency of acute respiratory distress syndrome (ARDS), sepsis, renal failure, and mortality increased from group A toward group C. The REE test revealed wide individual variation and was usually overestimated by all tested formulas. The mean REE was comparable in groups A, B, and C during the first 20 days (49 +/- 16% vs. 59 +/- 21% vs. 57 +/- 18% above the REE calculated by the Harris-Benedict equation, or HBEE). The REE of patients in groups A and B declined after this period, whereas the long-term ventilated patients in the prognostically unfavorable group C showed a high REE up to the 45th day, usually accompanied by severe organ dysfunction and major metabolic disorders. During this time a nutritional regimen meeting the actual REE could not be achieved. In the clinical situation when indirect calorimetry is not available, REE can be stated to be 50% to 60% above HBEE in patients with >20% BSAB for at least 20 days. Expecting a stable clinical course in patients with a predicted mortality of <20% (group A), oral nutrition usually seems sufficient after a short period of artificial nutritional support (1 week). Patients with a predicted mortality of more than 20% have a complication-burdened clinical course and a prolonged period of ventilation (groups B and C). These patients need parenteral and enteral nutrition for at least 20 days after trauma to prevent severe malnutrition.