BACKGROUND: Low anterior resection with coloanal anastomosis prevents a definitive stoma in patients with distal rectal cancer. However, imperative stool urge, stool fragmentation, prolonged stooling sessions, and minor problems of incontinence are frequently observed in the postoperative situation and negatively affect quality of life. Therefore, the colonic J-pouch was originally constructed to create a stool reservoir. In a randomized, prospective study, the short (5 cm) colonic J-pouch was tested for function and continence vs. straight coloanal anastomosis. METHODS: Over a period of 30 months, 74 consecutive patients (55 males) with rectal cancer in the lower and middle third of the rectum were included and randomized into two groups. Anastomosis was performed either as a coloanal or a colon-pouch-anal anastomosis. The standardized surgical procedure included mobilization of the left hemicolon, central ligation of the inferior mesenteric artery and vein, preaortal lymph node dissection, autonomic nerve preservation, and total mesorectal excision. The anastomosis was performed at the upper anal canal or at the intersphincteric level. All patients were evaluated preoperatively and six months postoperatively for fecal continence, including sphincter manometry and defecation habits. In addition, quality of life was determined by use of a standardized questionnaire (European Organization for Research and Treatment of Cancer, EORTC-QLQ-C30). RESULTS: Thirty-seven patients were randomized into each group. In general, problems with continence for liquids or gas occurred less frequently in the colonic J-pouch group 6 months after surgery. The frequency of bowel movements was lower in the J-pouch group (2.5 per day) than in the coloanal group (4.7 per day). Importantly, in a manometric study at the same postoperative point, neorectal capacity was decreased to a similar degree in both groups compared with the preoperative rectal volume. Thus, the expected and postulated reservoir effect could not be achieved by forming a 5-cm colonic J-pouch. CONCLUSION: The colonic J-pouch was superior with regard to continence for gas and liquids compared with a straight coloanal anastomosis. Furthermore, stool frequency was significantly lower in the J-pouch group than in the coloanal reconstruction group. However, because neorectal capacity decreased equally in both groups, we speculate that the advantage of the colonic J-pouch is not in the creation of a larger neorectal reservoir but rather may be related to decreased motility.