BACKGROUND Survival rates after surgery for rectal cancer remain at about 40% at 5 years from diagnosis. The aim of this study was to find out whether local recurrence rate could be reduced and survival increased by a moderately high dose of preoperative radiotherapy in patients with locally advanced, but otherwise operable, carcinoma of the rectum. METHODS We carried out a prospective randomised trial of surgery alone (n = 140) versus surgery preceded by 40 Gy radiotherapy (n = 139) given in 20 fractions of 2 Gy over 4 weeks. The patients, from 20 regional centres throughout the UK, were enrolled between 1981 and 1989, and followed up for a minimum of 5 years or to death. FINDINGS 217 patients died, 114 of 140 allocated surgery alone and 103 of 139 allocated preoperative radiotherapy: median survival times were 24 months and 31 months, respectively. The hazard ratio for overall survival was 0.79 (95% CI 0.60-1.04, p = 0.10). At 5 years' follow-up 65 patients allocated surgery alone and 50 who received preoperative radiotherapy had local recurrence (hazard ratio 0.68 [0.47-0.98], p = 0.04); the corresponding numbers of patients with distant recurrence were 67 and 49 (hazard ratio 0.66 [0.46-0.95], p = 0.02). There was a significant benefit of radiotherapy on disease-free survival (hazard ratio 0.76 [0.58-1.0], p = 0.05). There was no increase in postoperative or late complications in the radiotherapy group. INTERPRETATION Our results provide further evidence that preoperative radiotherapy can reduce the rate of local recurrence of rectal cancer in patients with locally advanced disease. However, survival results are still equivocal, and so we must await the results of a meta-analysis of all radiotherapy trials from which precise and definitive results, particularly for survival, may be obtained.