This study was designed to determine functional outcomes of restorative resections for carcinoma of the rectum. Between 1992 and 1995, 16 patients (8 male) of median age 75 (range 58-88) years underwent resection and coloanal anastomosis with J-colonic pouch reconstruction for rectal cancer. The distance of the lower border of tumor from the anal verge was 5 cm in 7 patients (43.70-0), 6 to 7 cm in 6 patients (37.0%) and 8 to 10 cm in 3 patients (18.7%). The Dukes staging was A in 4 patients (25%), B in 8 patients (50%) and C in 4 patients (25%). Coloanal anastomosis was performed by hand in 4 patients (25%) or with stapler in 12 patients (75%). Seven patients (43.7%) had diverting stoma. No patient died following surgery. Pelvic sepsis and anastomotic dehiscence that required pouch excision occurred in two patients (12.5%). Fourteen patients were evaluated. The mean follow-up was 15 (range 5-23) months. No patient developed recurrent pelvic tumor. Satisfactory fecal continence was achieved by 85.7% of patients. Stool frequency was 1-2 per day in 12 patients (85.7%). Three patients (21.4%) had minor soiling. Five patients (35.7%) occasionally complained urgency. There was no apparent difference between the patients with manual or stapled coloanal anastomosis. Coloanal anastomosis with J-colonic pouch is associated with an excellent long-term outcome.