A randomized, multicenter clinical trial was conducted in Western Norway to study the effectiveness of preoperative radiation therapy in operable rectal cancer, given at a dosage of 3150 cGy in 18 fractions, 2 to 3 weeks before radical surgery. Three hundred nine patients were entered into the trial between May 1976 and December 1985. After radiation no tumor was seen in 4.5% of the patients. There was no increased morbidity or mortality at surgery. The 5-year survival for evaluable patients was 57.5% in the control group and 56.7% in the radiotherapy group. For patients operated on for cure the 5-year survival was 60.9% and 64.2% in the control group and radiotherapy group, respectively. Radiation significantly delayed both local and distant recurrences in patients in the radiation group who had curative resection from 13.3 months in controls to 27.1 months. The local recurrence rate in the corresponding groups was 21.1% and 13.7%, respectively. We conclude that higher preoperative radiation doses should be used in new trials as a higher dosage may transform the observed positive effects into a survival benefit.