A randomized, prospective and double blind trial was conducted on 308 patients. Thirty-one out of 102 patients receiving metronidazole developed wound infection compared with 13 out of 103 patients receiving cefotaxime and 14 out of 103 patients receiving metronidazole and cefotaxime. The reduction in the wound sepsis by the use of cefotaxime or a combination of metronidazole and cefotaxime was statistically significant when compared with patients who received metronidazole only. For early cases including normal, acutely inflamed and gangrenous appendices, we gave a single dose of antibiotics just before operation. For late cases with perforation and abscess formation, the antibiotics were continued for 5 days. There was no statistical difference in wound infection between patients who received cefotaxime alone and these who received a combination of metronidazole and cefotaxime.