systems, the energy used in the synthesis of body componentes and also energy expended to maintain body temperature.The energy expenditure at rest can be measured by Icorestimatedby predictive equations. For instance, it is said that the Harris-Benedict equation, which is the most used, estimates the resting energy expenditure with an accuracy of ± 10% in 80% to 90% of normal individuals. When used in critically ill patients, however, this equation correctly predicts the expenditure in less than 50% of subjects. This variability is related to the severity of the disease, changes in body composition, altered metabolic activity and multiple thermogenic clinical interventions [1].Recognizing the need to estimate energy expenditure in institutions that do not have IC, several researchers have proposed specific equations developed from calorimetry studies in groups of patients with similar clinical characteristics.A study in healthy Brazilians living in the southwestern region of the United States, in order to estimate the accuracy of some predictive equations to estimate the BME, showed that the Schofield equation over estimated the expenditure measured by calorimetry at about 8.5 to 17, 5% and Harris -Benedict over estimated calorimetry data at 15 to 16.8%. They conclude that these equations, recommended for international use, were not suitable for use in the studied Brazilians [2].A Dutch group also studied adult patients comparing the BME data obtained by indirect calorimetry with different predictive equations based on weight, height, genderand age. Eighteen equations were studied, including 48 outpatients and 45 inpatients. The percentage of patients with acceptable predictions was only 25-52%, depending on the equation used. The best equation was recommended by the FAO / WHO / UNU both in hospitalized patients and out patients, and that is the equation recommended by the group when you do not have calorimetry [3].In another scenario (obese adults hospitalized with body mass index greaterthan 30 kg/m2), researchers compared the BME data obtained by calorimetry with several predictive equations, including the Harris-Benedict, Mifflin, Ireton Jones and others. Only the Harris -Benedict equationes timated BME within 10% error compared to the indirect calorimetry, in only 50% of patients [4].However a trial involving 337 individuals of a community in the US, studied by indirect calorimetry and comparisons with the Miffl in StJeor, Harris Benedict, Livingston, Muller and other equations, suggested that the Mifflin StJeor and Livingston equations are useful to predict the BME in obese and non-obese adults, while the other equations over estimated BEM [5].A Brazilian study attempted to standardize an equation to be used specifically for Brazilian women. Seven hundredand sixty women between 18 and 65 years were assessed by indirect calorimetry and the data correlated by regression logistics to age, weight and height. Two formulas were obtained, one for women with BMI above 35 kg / m2, and one for women below this score...