Various factors may prolong postoperative recovery and the length of stay (LOS) in the hospital. In a retrospective study of 245 adult patients who had no oral intake for 5 days after major gastrointestinal surgery, we used a correlation matrix to describe the population and determine the effects of the following factors on LOS: malnutrition, complication status, stress level, type of surgery, pathology, period of inadequate nutrient intake, and use of nutritional support. LOS was markedly prolonged in malnourished patients compared with those who were not (23.5 +/- 16.5 vs. 16.5 +/- 10.7 days, means +/- SD, p < 0.001). Patients were then grouped into those who had nutritional support and those who had not, and a nutrition classification was derived by examining the uncertainty (entropy) in the data matrix that allowed separation of the population into distinct groups. Nutrition and complication status and days without oral nutrient intake were discriminative. Analysis of variance and multivariate studies were also used to determine whether the presence of malnutrition, complications, both together, or neither could predict LOS and to determine the confounding effect of nutritional support on LOS. A significantly extended LOS persisted for patients with malnutrition or complications and was most prolonged for those with both, but patients who received nutritional support had a greater LOS than those who did not. In addition to the effects of nutritional support, malnutrition, and complication status, LOS correlated with the duration of the postoperative period without oral nutrient intake. We therefore recommend systematic and early nutritional intervention for selected gastrointestinal surgical patients.