A randomized prospective study of the relative effectiveness of broad-spectrum versus specific antistaphylococcal antibiotic prophylaxis in open-heart surgery was done between May 1972 and June 1973. All patients undergoing open-heart surgery were assigned randomly (by hospital number) to receive either methicillin (M) or cephalothin (C) beginning the night before operation. There were 132 patients in the C group and 129 in the M group. There were no statistically significant differences in age or duration of hospitalization, cardiopulmonary bypass, urinary tract drainage, or postoperative fever. There was a significant difference in the ratio of male to total patients (0.67 C, 0.52 M, P less than .02) and duration of operation in hours (4.27 C, 3.87 M, P less than .02). A statistically significant higher rate of urinary tract infection (3 C, 22 M, P less than .05), pneumonia (0 C, 9 M, P less than .01), and episodes of sepsis and prosthetic endo carditis (0 C, 11 M, P less than .001) was found in the methicillin group. The incidence of wound infections and positive cultures from blood obtained immediately after termination of cardiopulmonary bypass was not significantly different. Cephalothin replaced methicillin as the routine antibiotic prophylaxis for open-heart surgery at our institution.