OBJECTIVE: To assess the results of standardised total mesorectal excision of rectal cancer with particular reference to local recurrence and survival. DESIGN: Prospective open study. SETTING: Central hospital, Norway. MAIN OUTCOME MEASURES: Local recurrence, survival. RESULTS: The resectability rate was 90% (107/118), of whom 81 (76%) underwent curative resection. The overall local recurrence rate in patients who underwent primary resection was 9% with an overall five-year survival rate of 53%. In patients who had had curative operations the local recurrence rate was 4% (3/81), with an overall five year survival of 65% and a cancer specific survival of 85%. None of the patients who had palliative treatment survived five years. In 12 patients whose tumours were thought to be unresectable but who were operated on, of whom nine were given additional radiotherapy (46 Gy), 5 (42%) developed local recurrences and the five year cancer free survival was 25%. CONCLUSION: Total mesorectal excision and strict adherence to the surgical principles of anatomical dissection in the pelvis and washing out of the rectal stump before anastomosis reduce local recurrences to a minimum. In patients with locally advanced, fixed cancers, preoperative irradiation with more than 46 Gy must precede operation to achieve local control.