Between February 1981 and September 1989, 144 patients with rectal cancer were treated according to a radiosurgical sandwich protocol (27Gy + surgery + 18Gy) or postoperative radiotherapy (45Gy) at the University Hospital 'A. Gemelli' in Rome. This group is compared with a group of 133 patients operated on between January 1968 and January 1981, by the same team of surgeons but who received no radiotherapy. The historical group is comparable to the radiosurgical group in terms of stage, histology, and surgical procedures. The median follow-up period of the radiosurgical group is 38 months and 68% of cases have been observed longer than 2 years. At 2 years local recurrence in the historical control group was 22% versus 17% in the prospective group (p = 0.8). For Stage C disease, local recurrence dropped from 54% to 35% with adjuvant radiotherapy (p = 0.3). Metastases were observed in 22% of the control group versus 13% of the radiosurgical group (p = 0.2). For Stage C disease the incidence of distant metastases dropped from 59% to 26% at 2 years with the use of the prospective radiosurgical protocol (p = 0.05). The Kaplan-Meier survival rate at 5 years was 46% for the historical group and 72% for the radiosurgical group (p = 0.003) (Stage A 71% & 94%, Stage B 48% & 77%, and Stage C 16% & 38%, respectively). Neither serious nor late toxicity has been detected in the radiotherapy group, nor were surgical complications observed in the pre-operative radiotherapy group. The data strongly suggest a survival advantage for patients treated with the radiosurgical combination (p = 0.003).