We compared the effects of single dose (750 milligrams) prophylactic cefamandole delivered directly into the operative wound with local anesthesia (n = 162) with a control group (no antibiotics) (n = 162) in a randomized trial. No adverse effects were observed. There were seven wound abscesses in the untreated group compared with none in the group receiving antibiotic prophylaxis (p = 0.007). Six of the seven abscesses occurred as late as one month after the patient was discharged from the hospital. The costs of antibiotics used were ten times less than the costs of treatment of wound complications in the control group.