Bile acids are known to promote colon carcinogenesis. However, there is one study showing that ursodeoxycholic acid (UDCA) supplemented in the diet at the concentration of 0.4% prevented azoxymethane-induced rat colon tumorigenesis. The aim of our study was to explore the inhibitory effect of a much smaller dose of UDCA on colon carcinogenesis in rats. One hundred 7-week-old F344 rats were given 2 mg of N-methylnitrosourea 3 times a week for 3 weeks by intrarectal instillation, and were fed a 0% (control), 0.4% or 0.08% UDCA-supplemented diet for the next 27 weeks. All the rats were killed and examined for tumor development at week 30. The tumor incidence and number were significantly lower and smaller, respectively, in the UDCA-fed rats than in the control rats: 40% and 36% vs. 68%; 0.5 +/- 0.1 (mean +/- SEM) and 0.4 +/- 0.1 vs. 1.0 +/- 0.2. All the tumors were located in the distal half of the colon and were plaque-shaped or polypoid, being well-differentiated adenocarcinomas restricted to the mucosa or submucosa. Bile acids in the feces and the blood obtained at weeks 20 and 30, respectively, were analyzed by HPLC. A significant increase of UDCA was confirmed in both the feces and the blood of the UDCA-fed rats compared with the control rats. The results suggest that the continuous feeding of a small dose of UDCA may prevent colon carcinogenesis.