&NA;The mechanism by which ureterocolic anastomoses promote cancer of the colon is uncertain. To investigate the role of anastomotic healing, rats were exposed to the colon carcinogen azoxymethane after which they had either 1) a sham operation, 2) colotomy with suture, or 3) colovesical anastomosis, performed randomly. The first two groups had an equally low frequency of colon cancers, whereas rats with colovesical fistulas had a significantly higher frequency, and cancers were concentrated at the anastomotic site. It was concluded that prompt healing does not promote cancer in this model. It is probably not responsible for suture‐line recurrences seen clinically, nor for the increased frequency of cancer at the site of urinary diversion into the colon.