The duration of antimicrobial prophylaxis in gastric cancer surgery is not yet established. This prospective randomized study was performed to confirm the noninferiority of single-dose versus multiple-dose antimicrobial prophylaxis in terms of the incidence of surgical-site infection in gastric cancer surgery. Three hundred twenty-five patients undergoing elective resection for gastric cancer were randomized to receive only single-dose cefazolin (1 g) during surgery (single-dose group) or an additional 5 doses every 12 hours postoperatively (multiple-dose group). The overall incidence of surgical-site infections was 9.1% in the single-dose group and 6.2% in the multiple-dose group [difference (95% confidence interval): -2.9% (-5.9%-0.0%)]. Multivariate logistic regression analysis identified blood loss, being overweight, and advanced age as significant independent risk factors for surgical-site infection. Single-dose antimicrobial prophylaxis seemed to be acceptable, and choosing multiple-dose prophylaxis may have little impact on the prevention of surgical-site infections in elective gastric cancer surgery.