BACKGROUND A number of patients suffer from gastrointestinal side effects following sphincter-saving resection of the rectum. The aim of this study was to assess frequency and risk factors of long-term gastrointestinal side effects after sphincter-saving resection for rectal cancer. METHODS Between 1980 and 1997, among 209 patients treated for rectal cancer by rectal resection and sphincter conservation, 145 who were alive without recurrence or colostomy, responded to a questionnaire. There were 85 males and 60 females with a mean age of 64 Years; the follow-up was 5.4 Years. The functional result was considered good if the number of stools per 24h was three or less with normal continence and poor if the number of stools was four or more or if fecal incontinence was present. Influence of age, gender, anastomotic height, type of sutured colon, colonic pouch, defunctioning stoma, leakage, stenosis, radiotherapy, history of irritable bowel syndrome and follow-up were analyzed. RESULTS The mean anastomotic height was 5 cm from the anal margin. Functional results were good in 56% of patients and poor in 44%. Univariate analysis showed three variables to be significantly associated with the functional results: anastomotic height (P=0.001), radiotherapy (P=0.03) and follow-up > 24 Months (P=0.009). Multivariate analysis showed that only anastomotic height independently influenced functional results. They were good in 76%, 53% and 35% of patients for anastomoses located above 6 cm, between 6 and 3 cm, and below 3 cm from the anal margin, respectively. CONCLUSIONS After sphincter-saving resection for rectal cancer, about half of patients have poor long-term functional results, the latter being related only to the anastomotic height. This suggests the need to preserve a small part of the rectum when oncologically feasible.