OBJECTIVE To compare metabolic, nutritional and epidemiological data in two groups of patients, one receiving total enteral nutrition, via nasoenteric tube, and one receiving both enteral and parenteral nutrition. METHODS A prospective, randomized study. METHODS A general ICU, with both medical and surgical patients, in a big regional University and National Health Service hospital. METHODS 24 patients requiring Intensive Care after major surgery or because suffering from severe head injury or major neurological impairment. METHODS All patients initially received total parenteral nutrition: after 4 days 12 patients were "weaned" to total enteral nutrition and 12 stayed on mixed parenteral and enteral nutrition. LABORATORY INVESTIGATIONS AND OBSERVATIONAL DATA: Blood levels of albumin, prealbumin, transferrin, ALT, AST, bilirubin, blood urea, blood glucose, total linfocite count, and nutritional and epidemiological data such as nitrogen balance, calorie intake, diarrhea incidence, blood and sputum cultures and radiologic evidence of pneumonia are analysed. RESULTS At T1, NET patients were able to reduce their nitrogen losses (0.27.1 g/kg +/-0.12 vs 0.35 +/- 0.13 at TO; p < 0.05) and improve nitrogen balance (-9 +/- 7 vs -2 +/- 6 at T0; p < 0.05); they also had a better total linfocite count (2034 +/- 304 vs. 1413 +/- 360 of the MISTA group; p < 0.05), and a lower incidence of pneumonia as documented by sputum cultures and radiograms. CONCLUSIONS Patients fed with both parenteral and enteral nutrition did no better than those on total enteral nutrition as far as nutritional and metabolic indices were concerned; they also seemed more prone to infections than those on total enteral nutrition, indicating that mixed nutrition may result in more stable feeding, but this does not seem to have any beneficial nutritional, immunological and metabolic effect.