In a prospective, randomized trial involving 100 patients with severe intra-abdominal sepsis, the value of single agent antibiotic therapy with cefotetan was compared with that of combination therapy of ampicillin, gentamicin and metronidazole (AGM). All patients underwent exploratory laparotomy. The mortality rate was 3 per cent, all deaths occurring within 48 h of operation. Two-thirds of patients were considered severely ill on admission, and one-third were moderately ill. Six patients had positive blood cultures on entry into the study. The mean age was 31 years and concurrent disease was present in 14 per cent of the patients. A satisfactory response was achieved in 82 per cent of patients receiving cefotetan and in 65 per cent of those receiving AGM, whereas the response was unsatisfactory in 18 per cent of cefotetan patients and 35 per cent of those receiving AGM (P = 0.075 n.s.). Significant changes in laboratory values during the study occurred in 51 per cent of patients, and 7 per cent required vitamin K administration for hypoprothrombinaemia. The results of this study suggest that antibiotic therapy with single agent cefotetan is as safe and effective as a combination of ampicillin, gentamicin and metronidazole in patients with severe intra-abdominal sepsis requiring operative management.