In order to obtain information on the preventive effects of various food pro-teins against colonic cancer, six groups of azoxymethane-initiated mature Fischer rats (n=10) were fed respective diets different in protein sources such as bovine milk casein (ca-sein), high-molecular-weight fraction from protolytic digest of soy protein isolate (soybean HMF), hen's yolk defatted protein (yolk protein), wheat gluten and codfish meat, which had been supplemented with sodium deoxycholate (hereinafter, DCA) as a cancer promoter ex-cept for an additional DCA-unfed casein group. All of the living rats at checkpoints during the feeding period were examined by the use of a bronchus fiberscope for colonic tumor inci-dence at 6 wk intervals between the 10th and 34th wk, from which both blood and feces samples were taken at times of endoscopy. Tumorigenesis in the colon was perceived by en-doscopy at wk 22 in the group fed DCA casein only and at wk 28 in the other groups except the DCA-unfed casein group. At wk 34, both soybean HMF and yolk protein groups ranked inferior to the DCA-unfed group in tumor incidence. When plasma steroid or lipid concen-tration was plotted against tumor incidence at wk 2 8 or 34, positive correlations were found between plasma bile acid concentration and tumor incidence at both weeks. With the exception of the DCA-unfed casein group, plasma bile acid concentration was reversely cor-related to fecal bile acid excretion. Taken altogether, these results suggest that bile acids at higher concentrations in the plasma may serve as risk factors of colon tumor incidence.