From January, 1985, to June, 1993, 125 patients with stages B2-C adenocarcinomas of the rectum were submitted to pre- and postoperative irradiation according to Thomas Jefferson University protocol guidelines. Five hundred cGy were administered as a single preoperative dose 24 hours before surgery using parallel opposed (AP-PA) treatment fields including the whole pelvis. Pathologic samples were classified following the Astler-Coller staging criteria. Forty-seven patients had no postoperative treatment because their disease stage was A, B1 or D, 11 for refused consent and 9 postoperative complications preventing any further therapy. Seventy-eight patients concluded the treatment schedule and are assessable for response. Radiotherapy total dose consisted of 4400-5000 cGy administered over 5-6 weeks: the patients were treated with megavoltage photons (15-MeV photons) and one dose fraction of 2 Gy was delivered daily, 5 days a week, with the "box" or the "three-field" technique. Median follow-up time was 50.2 months from the beginning of treatment for all the patients in our series (range: 18-120 months). Radiation therapy was well tolerated: 5 patients had severe diarrhea and 2 had small bowel obstruction which required surgery. Local recurrences were observed in 13 of 78 patients (16.7%). Overall actuarial survival at 5 years was 66.8%. Our results confirm the efficacy of this treatment, which is in agreement with international literature data. However, no difference was seen relative to the results obtained with postoperative irradiation alone. We conclude that sandwich radiotherapy can be an effective tool for the local control of rectal adenocarcinoma, with acceptable morbidity, even though it fails to prevent metastases.