High energy deficits due to underfeeding are frequently observed during critical illness and are associated with significant morbidity and mortality. Adequate determination of energy requirements is imperative for optimizing nutrition. For this goal, indirect calorimetry is considered to be the gold standard but it is expensive, time-consuming, and not readily available in many hospitals. As an alternative, most ICU physicians use bedside formulas to predict calorie needs. Some equations are obtained during resting metabolism in healthy humans and “corrected” by adding stress or injury factors. Others are derived from regression analysis in patients whereby various static and dynamic variables are identified and eventually adjusted for type of patient and/or disease. Few studies have evaluated the accuracy of predictive equations in critically ill patients. The largest prospective study to date identified the Penn State equation, including a modified version for obesity, as being the most accurate. Whether the systematic use of (a) particular formula(s) for estimating calorie needs may influence morbidity or outcome in ICU patients remains to be determined.