The results of surgical treatment of carcinoma of the rectum have been improved in the last decades, because of the availability of stapling device, the increasing use of adjuvant or neoadjuvant treatment and the development of endoluminal surgery. Herein we present our thirty years experience in the treatment of rectal carcinoma. It deals with 690 patients operated on, 602 with conventional surgery (87%) and 88 with local excision (13%). As regards the first subgroup, in 63 patients (10.4%) the operation was considered palliative, because there were either hepatic metastases or residual tumor; 21% of the whole number had tumor in the upper third, 40% in the middle third and 39% in the lower third. Radical resection was performed in 538 patients (89.6%); 53% of the patients underwent anterior resection and 47% underwent abdomino-perineal resection. Ninety percent of tumors located in the lower rectum were managed with abdomino-perineal resection and 97% of tumors located in the upper rectum were managed with anterior resection. Tumor of the middle rectum were treated indifferently with anterior resection or abdomino-perineal resection. Mortality was 4%; the morbidity was 33%. No statistical difference was found between anterior resection and Miles operation. Mean follow up is 6.8 years. Local recurrence was observed in 17% of patients and was more frequent after Miles operation than after anterior resection. The 5- and 10-year survival is respectively 53% and 43%; according to staging, 5-year survival is 68% for stage I, 64% for stage II, 39% for stage III and 9% for stage IV (p < 0.0001). As regard local excision, lesions usually arised from lower or middle rectum and were less than 3 cm. in the maximum diameter. Sixty per cent of patients received adjuvant or neoadjuvant treatment; we observed complete response with disappearance of the tumor in 28% of T2 patients. The toxicity was low: 3% of patients developed erithema and 21% different grades of proctitis. Mean duration of operation was 179 +/- 98 min' and hospitalization was less than 7 days in most of the patients. No mortality was recorded, while morbidity was 16% (14 patients). On pathological examination we observed 10 Tis (11%), 32 T1 (36%)28 T2 (32%) and 18 T3 (21%). Mean follow up is 3.3 years; local recurrence was related to staging and was respectively 0%, 16%, 25% and 61% (p = 0.0008). We did not observe local recurrence whenever T2 patients had neoadjuvant treatment and negative margins at i.o. pathological examination. Our findings show that anterior resection represents the most appropriate treatment for tumors of the upper and middle rectum; Miles operation should be reserved to tumors located within 5 cm. from the anal verge. Small tumors, confined to the rectal wall, can be properly treated with local excision; local treatment should be always associated to preoperative chemoradiation in tumors involving the muscular layer of the rectal wall.