Continence-preserving coloproctectomy for ulcerative colitis is technically demanding and is relatively often afflicted with complications. We have retrospectively reviewed the files of all patients being operated for ulcerative colitis at Ullevål Hospital from 1992 to 1997 (n = 53). Most of the patients (n = 50) were examined clinically; 12 patients had anal manometry before and after operation. 44 patients were operated with continence-preserving coloproctectomy with J-pouch and handsewn anastomosis; of these, 42 were followed more than six months. Eight had pouchitis, four perianal abscess/fistula, three septicaemia and three were operated for ileus. Two had anastomotic leakage and pelvic abscess that required transanal drainage. One had the pouch removed six years after operation due to chronic pouchitis and pouch-vaginal fistula. There was no deterioration of anal maximal resting and squeezing pressures on pre- and post-operative anal manometry. Mean number of stools from the reservoir per 24 hours were 6.2 (range 3-11); 11 patients had leakage of air and loose stool, three at day-time and eight at night. Two patients (4%) died from colorectal cancer and three (7%) had Crohn's disease. Nine patients were unfit for pouch surgery and underwent coloproctectomy (n = 7) or subtotal colectomy (n = 2). Our results indicate that pouch surgery for ulcerative colitis is a good option for most patients.