Two hundred fourteen women having vaginal or abdominal hysterectomy were entered into a prospective, randomized, blind clinical trial comparing a preoperative intravenous dose of piperacillin to three perioperative intravenous doses of cefoxitin given over an eight-hour period. Interregimen clinical, surgical, and outcome variables of the 207 evaluable subjects were statistically similar, but there were significant interprocedure differences in a variety of categories; many benefits exist when vaginal hysterectomy is possible. Efficacy of a single dose of piperacillin was similar to that of three cefoxitin doses. Seven women (3.4%) had major postoperative infection requiring parenteral antimicrobial therapy, two (1.9%) after vaginal hysterectomy and five (4.8%) after abdominal hysterectomy. Three of the latter five infections (60%) occurred after discharge from the hospital. Even with prophylaxis, postoperative anemia was associated with increased frequency of infection at the operative site after both procedures, and diabetes was associated with late infection of the abdominal incision after abdominal hysterectomy.