BACKGROUND: Anastomotic leakage occurs after 3-6 per cent of colonic resections. The influence of analgesic agents is largely unknown. This study determined the rate of anastomotic leakage in a series of patients who had colonic surgery over a 9-year period with or without use of a cyclo-oxygenase inhibitor for postoperative analgesia. METHODS: Patients with anastomotic leakage following a standard fast-track procotol between April 1997 and May 2006 were identified from a prospective, consecutive database. During this period there were two changes in perioperative management: cessation of preoperative oral bowel preparation in August 2002 and the use of celecoxib for postoperative analgesia between May 2003 and November 2004. Rates of anastomotic leakage during the various periods were determined and compared. RESULTS: Some 28 (5.6 per cent) of 502 patients had an anastomotic leak. The incidence of leakage increased significantly during the period of celecoxib use (15.1 per cent), versus 3.3 and 1.5 per cent respectively before and after celecoxib use (P < 0.001). Leakage rates were similar with or without oral bowel preparation (3.5 versus 1.7 per cent respectively; P = 0.346) when celecoxib was not used. CONCLUSION: A detrimental effect of celecoxib on anastomotic healing is suggested, and requires further evaluation.