The authors report their results of a prospective study on 1182 patients who underwent surgical operation relatively on postoperative infections. Studied variables were: structural and anamnestic: sex, age, smoking, drinking; clinical: evidence of functional changes in various organs, as assessed upon clinical basis and laboratory results; pertinent to surgical intervention: entity, duration, anaesthesia; during and early-after-surgery complications (until discharge or within 30 days since intervention). Stepwise logistic regression model was applied to this set of preoperative and operative factors, four of which were found significantly correlate with postoperative infections: bacterial contamination during surgery, duration surgical intervention greater than 120', cholestasis, serum albumin.