Antibioprophylaxis has been proven to be efficient for some orthopaedic procedures. However, its efficacy for clean limited procedures with metallic implants is not clear. One hundred twenty-two closed ankle fracture patients undergoing open reduction and internal fixation were randomized, double-blindly, to receive either cephalotin (1 g i.v. every 6 h x four doses) or a placebo starting before tourniquet application. Mean trauma-surgery delay was 30.2 h, and average tourniquet duration was 65.4 min. Four patients (three of 62 on placebo and one of 60 on cephalotin) developed a superficial wound infection. One of the patients in the placebo group was hospitalized and treated with i.v. antibiotics. However, there was no osteomyelitis or premature hardware removal. The difference between the two groups was not statistically significant (x2 test of appreciation p = 0.33, two-tailed probabilities). Therefore, cephalotin prophylaxis does not seem justified in this patient population. A larger series is needed to avoid a type II error.