300 patients undergoing various operations were randomly allocated to receive, intra-incisionally, just before skin suture, either a solution of 1 g cephaloridine or Polybactrin aerosol. An equal number of patients were allocated to each group, and various factors likely to influence the incidence of wound infections were evenly distributed. 13 wound infections (9.8%) developed in the cephaloridine group and 28 (20.1%) in the Polybactrin group (p < 0.02). In the former group, the infections were milder and there were 44 fewer patient-days of hospitalisation. The difference in infection rates for ‘potentially contaminated and contaminated’ operations was greater (p < 0.01), in favour of cephaloridine. Other differences are discussed and some reasons for the superiority of cephaloridine over other agents (e.g., povidone-iodine, ampicillin) are advan