A prospective randomized study was performed with 65 patients undergoing elective surgery for colorectal cancer, to evaluate the prophylactic effect of two different parenteral antibiotic regimens. All patients underwent rigorous mechanical cleansing of the bowel (enemas, laxatives), received low-residue diet 3 days pre-operatively, and were given oral metronidazole (250 mg) five times a day for 3 days preoperatively. They were divided into two groups comparable in age, nutritional status and operative procedure. The patients in group A (36) received 2 g i.v. of cefotetan at induction of anaesthesia and two other administrations every 12 h. Patients in group B (29) were given clindamicin (600 mg, i.v.) at induction of anaesthesia plus aztreonam (1 g, i.v.); two other doses of the same combined antibiotics were administered every 8 h. Five patients were excluded from the study because they underwent Miles procedure; two others because they underwent explorative laparotomy only. The overall incidence of post-operative septic complications was 6.9% (4/58). No significant differences were found in terms of the rate of surgical infections: 3.1 in group A (1/32) and 0% in group B. Urinary tract infections (1 case) and respiratory tract infections (2 cases) were observed only in group B: the rate was found to be 11.5% (3/26); two anastomotic leakages were observed in group A (6.25%) and one in group B (3.8%). These data suggest that cefotetan appears to be as effective as clindamicin plus aztreonam in prophylaxis against infection in elective colorectal surgery.