Laboratory results indicate that the endocervix may be a source of bacterial contamination when vaginal hysterectomy is performed. In a series of 160 consecutive vaginal hysterectomies in premenopausal women, hot conization of the cervix was performed prior to the scrub with an iodophore. No preoperative antibiotics were used in this series. The postoperative febrile morbidity rate was 4.3 per cent and the average stay was 4.5 days. These results are compared with those of three other groups: (1) patients who received a three-dose parenteral prophylactic antibiotic course with the first dose two hours prior to surgery had a febrile morbidity rate of 8.6 per cent. (2) In patients who had prophylactic antibiotics for five days with the first dose given intraoperatively, the febrile morbidity rate was 10.1 per cent. (3) The febrile morbidity rate in the group with no antibiotic prophylaxis or hot conization was 49.1 per cent. Laboratory and clinical data suggest that preoperative conization may be effective in the reduction of postoperative febrile morbidity.