This study included 207 patients in a double-blind, prospective randomized trial of cefazolin versus cefamandole single-dose prophylaxis by both intravenous (IV) and lavage routes in patients laboring with ruptured membranes who eventually had nonelective cesareans. No significant differences (P less than .05) were noted in the rates of infectious morbidity (cefazolin IV: six of 47, 13%; cefazolin lavage: nine of 59, 15%; cefamandole IV: six of 47, 13%; cefamandole lavage: six of 54, 11%). The incidence of operative complications in the patients developing endometritis (eight of 22, 36%) was significantly different (P less than .004) from that in the group who did not develop endometritis (21 of 185, 11%). Cefazolin by IV and lavage routes of administration appears to be as effective as cefamandole by either route of administration.