A prospective multicenter study of the management of large bowel cancer recorded the results in 4,500 patients in whom 2,056 have had an elective colorectal anastomoses. Of these patients, 15.8 percent had a synchronous covering stoma to protect the anastomoses. Although the anastomotic leak rate was high in patients with a stoma, no overall differences were observed in mortality between those patients who had a covering stoma and those patients who did not (7 percent and 6.1 percent, respectively). However, when surgical policies were analyzed, clinically large and statistically significant differences were found. Some surgeons frequently used a covering stoma for low anterior resection whereas others only rarely did so. The differences in anastomotic leak and mortality were 20 percent and 7.8 percent, and 8.4 percent and 3.6 percent, respectively. We conclude that all surgeons should know their own clinical and radiologic anastomotic leak rate. If and when this figure becomes low (less than 5 percent), the covering stomas will become necessary except for the very rare and difficult case.