The outcome of patients with rectal cancer treated by abdominoperineal or anterior resection, with or without preoperative radiotherapy, was assessed to detect any differences attributable to the operative method and interactions between radiotherapy and type of surgery. The study was based on 1292 patients included in two consecutive controlled randomized trials of preoperative radiotherapy in operable rectal carcinoma. The outcome was not related to surgical method. Radiotherapy increased postoperative mortality and complications and reduced local and distant recurrence, but had no effect on overall survival. Effects of radiotherapy were similar irrespective of the type of surgery, except that the increase in postoperative mortality in irradiated patients was greater in those treated with abdominoperineal resection. Sphincter-saving procedures appear to have no adverse effects on outcome of rectal cancer, but the optimum use of radiotherapy is still to be defined.