This prospective randomized controlled clinical trial compares the effects of early parenteral nutrition and traditional delayed enteral nutrition upon the outcome of head-injured patients. Thirty-eight head-injured patients were randomly assigned to receive total parenteral nutrition (TPN) or standard enteral nutrition (SEN). Clinical and nutritional data were collected on all patients until death or for 18 days of hospitalization. Survival and functional recovery were monitored in survivors for 1 year. Of the 38 patients, 18 were randomized to the SEN group and 20 to the TPN group. Demographically, the two groups of patients were similar on admission. There was no significant difference in the severity of head injury between the two groups as measured by the Glasgow Coma Scale (p = 0.52). The outcome for the two groups was quite different, with eight of the 18 SEN patients dying within 18 days of injury, whereas no patient in the TPN group died within this period (p less than 0.0001). The basis for the improved survival in the TPN patients appears to be improved nutrition. The TPN patients had a more positive nitrogen balance (p less than 0.06), and a higher serum albumin level and total lymphocyte count. More adequate nutritional status may have improved the patients' immunocompetence, resulting in decreased susceptibility to sepsis. The data from this study strongly support the favorable effect of early TPN on survival from head injury.