The purpose of this prospective, randomized, double-blind clinical trial was to compare the efficacy of 1-g doses of intravenous cefazolin with that of 1-g doses of intravenous cefotetan in preventing major operative site infections after elective abdominal hysterectomy. A major operative site infection requiring parenteral antimicrobial therapy developed in 46 (9%) of 511 evaluable women: 30 (11.6%) of 258 women given cefazolin prophylaxis and 16 (6.3%) of 253 women given cefotetan prophylaxis (relative risk, 1.84; 95% confidence interval, 1.03 to 3.29; P < .05). Risk factors for major operative site infection were younger age, lower postoperative hemoglobin concentration, and a proliferative endometrium. Ten (3.9%) of 258 women given cefazolin prophylaxis had a postoperative pelvic abscess; two of these women required additional surgical procedures, compared with two (0.8%) of 253 women given cefotetan prophylaxis who had an abscess but did not require surgery (relative risk, 4.9; 95% confidence interval, 1.09 to 22.16; P = .04). A greater number of infections and more serious infections occurred following cefazolin prophylaxis; this treatment resulted in 234 additional hospital days for administration of parenteral antimicrobial therapy.