Sixty-two patients were admitted to a prospective randomized controlled trial to investigate the influence of a prophylactic antibiotic, lincomycin, on anaerobic sepsis following bowel surgery. The incidence of postoperative sepsis was reduced from 45 to 18 per cent (P less than 0-025). Wound infections were reduced from 38 to 12 percent (P less than 0-05). Intra-abdominal or pelvic abscess occurred in 1 of the treated group compared with 3 controls. Septicaemia occurred after operation in 1 patient receiving lincomycin and in 3 of the controls; in 2 of the latter, pure growths of bacteroides were isolated from the blood cultures and 1 of these patients died. Although lincomycin had no influence on the number of patients who developed aerobic postoperative infections, there was a significant reduction in the incidence of sepsis due to bacteroides, which occurred in 10 of the control group compared with 1 in the lincomycin group (P less than 0-005). No patients developed complications attributable to lincomycin, such as pseudomembranous colitis. These data indicate that the genus Bacteroides are important pathogenic organisms and are responsible for postoperative morbidity. Furthermore, anaerobic sepsis can be reduced by appropriate prophylactic antibiotics.