A randomized prospective trial was undertaken of polydioxanone suture (PDS) versus conventional suture material in 98 patients undergoing anastomoses in the gastrointestinal tract. Nine patients died within 6 months of surgery, one of these being related to an anastomotic leak. All other patients were followed up for between one and three and a half years. In 57 colonic anastomoses, 30 were randomized to a single layer of 2/0 (BPC) interrupted PDS and 27 to a single of 2/0 (BPC) interrupted silk. Follow up sigmoidoscopy and barium enemas were used to confirm the clinical suspicion of 6 benign anastomotic strictures, 5 of which occurred in the PDS group. At this stage, the colonic arm of the trial was discontinued because the 19% stricture rate with PDS was deemed unacceptable. In a second limb of the study, patients were randomized to two layers of 2/0 (BPC) continuous PDS or 2/0 (BPC) continuous chromic catgut. There were no significant differences in 32 gastric or small bowel anastomoses and in particular, no anastomotic strictures were apparent. The reasons for the high rate of stricture formation when using PDS for large bowel anastomoses are unclear. However, it would seem to be a suitable alternative to chromic catgut when confined to the stomach and small intestine.