A prospective randomized trial was performed on 122 patients with perforated or gangrenous appendicitis to determine whether delayed primary wound closure lowered morbidity from wound infection. The 54 per cent wound infection rate with delayed primary closure was significantly inferior to the 18 per cent infection rate for primary closure with topical ampicillin powder (P = 0.0082), but not significantly different from the 37 per cent infection rate for primary closure alone. Furthermore, patients disliked delayed primary closure, their hospital stay was prolonged and 17 per cent of their wounds became contaminated with Staphylococci before being closed. Delayed primary wound closure should not be used in treating perforated and gangrenous appendicitis wounds.