Previous studies have demonstrated that short course perioperative antibiotic prophylaxis reduces septic morbidity after hysterectomy from up to 40% down to 10 to 15%. The residual morbidity is predominantly urinary tract infection (UTI) occurring 2 to 3 days after cessation of antibiotic. We hypothesised that surgery impairs urinary drainage for 3 to 4 days postoperatively and that prolonged prophylaxis was required to prevent all postoperative sepsis. Accordingly, 224 abdominal and 69 vaginal hysterectomy patients were randomly allocated to short course (2g cefotaxime IV at anaesthesia) or long course (2g cefotaxime at anaesthesia plus 7 doses of 1g 12-hourly) prophylaxis. UTI was effectively prevented by the long course (0.9 vs 18.4% short); both treatments effectively prevented wound and pelvic sepsis. We conclude that 4 days of postoperative antibiotic is required to prevent UTI.