From 1980 to 1985, 471 patients with resectable rectal and rectosigmoid cancer were randomly allocated to receive either preoperative short-term high-dose irradiation (25.5 Gy in one week) for all patients or prolonged postoperative radiotherapy (60 Gy in seven to eight weeks) only for patients with a Dukes B or C lesion. After a minimum follow-up of five years, the local recurrence rate was statistically significantly lower after preoperative than after postoperative radiotherapy (13 percent vs. 22 percent; P = 0.02). No difference in overall survival was noted (P = 0.5). To evaluate possible late side effects on the bowel, urinary bladder, or skin after surgery and additional preoperative or postoperative radiotherapy, all patients included in the randomized trial, together with 58 patients from a preceding pilot study with the same preoperative regimen, were studied in a prolonged follow-up program. The hospital files of all patients were re-examined. Of the patients who were carefully examined, 176 had a survival exceeding five years and 19 had a survival exceeding 10 years. Overall, 7 percent (33/464) either were operated upon or have had a radiologic diagnosis of small bowel obstruction: 14/255 (5 percent) after preoperative irradiation, 14/127 (11 percent) after postoperative irradiation, and 5/82 (6 percent) after surgery alone. The cumulative risk of developing a bowel obstruction was significantly increased after postoperative radiotherapy. Among the 98 patients alive after preoperative irradiation, significant morbidity from the bowel was noted in 11 patients, from the urinary bladder in two, and from the skin in six. In the postoperatively treated group of 34 patients, the bowel, urinary bladder, and skin morbidity were significant in five, two, and five patients, respectively. Corresponding morbidity in 44 nonirradiated patients was seen in five, one, and two patients, respectively. It is concluded that preoperative, short-term, high-dose radiotherapy decreases the local recurrence rate relative to postoperative radiotherapy, with no indications of increased late morbidity after a follow-up of 5 to 10 years.