Nutritional support is routine practice in critically ill patients and enteral feeding is preferred to the parenteral route. However this direct delivery of nutrients to the gut is potentially ineffective for a variety of reasons. We performed a prospective audit of 40 consecutive intensive care patients to determine whether enteral feeding met the nutritional requirements of our patients. The ideal requirements for each patient were calculated using the Harris-Benedict equation with an adjustment determined by the patient's diagnosis. We compared the amount of feed delivered with the daily requirements over a seven-day period Successful feeding was defined as the achievement of 90% of the ideal calorie requirement for two consecutive days. The mean calculated (+/- SD) energy requirement was 9,566 kJ (+/- 2,586). Patients received only 51% (SD 38) of their energy requirements throughout the study period. Only 10 patients (25%) were successfully fed for at least any two-day period in the seven days. Feeding was limited mainly by gastrointestinal dysfunction or by the need to fast the patient for medical, surgical and airway procedures. Success of feeding was not related to the use of sedative orparalysing agents and had no correlation with plasma albumin concentration. There was no difference in the volume of feed delivered to patients who survived or died. Prokinetic agents were used in 25 patients and in these patients there was a trend towards improved delivery of feed.