In order to assess the significance of malnutrition in determining surgical complications and the possibility of their reduction by preoperative nutritional support (PNS), a randomized controlled trial is being performed at our institution. The results relative to 100 patients who underwent major surgery for gastrointestinal disease, are presented here. In the treatment group 49 patients received 30 kcal/kg/day and 200 mg/kg/day of nitrogen for at least 7 days in the immediate preoperative period (nine patients were excluded from this group due to early surgery--seven cases; or refusal to accept PNS--two cases. Data analysis with their inclusion or exclusion showed similar results.) Fifty-one patients constituted the control group. The observed septic complication rate was, respectively, 30 and 35.3% (p:NS). When the analysis was restricted to the patients with abnormal instant nutritional assessment (INA), as defined by Seltzer et al (serum albumin less than 3.5 g/dl and/or total lymphocyte count less than 1500 cells/mm3), a statistically significant difference was observed in the incidence of sepsis between the two subgroups (21% vs 53.3%, p less than 0.05). Analogous results were obtained from the patients who underwent gastrectomy for gastric cancer: 16.7% of septic complications in the malnourished treated patients and 100% in the malnourished control ones (p less than 0.05). The occurrence of serious sepsis (sepsis score greater than or equal to 10, according to the scoring system developed by Elebute and Stoner) in the malnourished subgroups was 5.2% and 26.7%, respectively, (p = 0.09). The postoperative mortality rate was not significantly changed by the PNS (reduction from 3.9% to 2.5%, p:NS).(ABSTRACT TRUNCATED AT 250 WORDS)