The rationale for the use of prophylactic antibiotics in clean orthopaedic surgery was explored by means of an experimental model suited to the investigation of (1) antibiotic penetration and persistence in operative hematomas in bone and (2) the effect of cephaloridine administration on standard wound infections. Cephaloridine was found to penetrate hematomas in bone readily and to persist there in bacteriocidal concentrations. When cephaloridine was given preoperatively, all wounds were converted to bacteriological sterility and did not become clinically infected. When cephaloridine was begun six hours or later after contamination, infection occurred uniformly.