Bifidobacterium longum has been shown to afford protection against colon tumorigenesis. Lactulose, a keto analog of lactose, serves as a substrate for preferential growth of Bifidobacterium. It is not known whether feeding lactulose along with B. longum will have any advantage over feeding of B. longum alone. To test this combination effect, 61 male Fisher 344 weanling rats were divided into four groups of 15 rats each (16 in the control group) and assigned to one of the following four diets for 13 weeks: (i) AIN76A (control, C); (ii) C + 0.5% B. longum (C+Bl, containing 1 x 10(8) viable cells/g feed); (iii) C + 2.5% lactulose (C+L); (iv) C + 0.5% B. longum + 2.5% lactulose (C+Bl+L). All animals received a s.c. injection of azoxymethane at 16 mg/kg body wt at 7 and 8 weeks of age. Colons of 10 rats from each dietary group were analyzed for aberrant crypt foci (ACF), which are preneoplastic markers. Colonic mucosa and livers from five rats were analyzed for glutathione S-transferase (GST, a Phase II enzyme marker). Results indicate that feeding of lactulose and B. longum singly and in combination reduces the number of ACF (P = 0.0001) and the total number of aberrant crypts significantly (P = 0.0005). The total number of ACF in diets C, C+Bl, C+L and C+Bl+L were 187 +/- 9, 143 +/- 9, 145 +/- 11 and 97 +/- 11 respectively. There was no significant difference in weight gain among treatments. Colonic mucosal GST levels were significantly (P = 0.05) higher in the Bl and L groups compared with group C. Initially there was a mild diarrhea in lactulose-fed rats. There was a positive correlation between higher cecal pH and number of ACF. Results of the study indicate that Bifidobacterium and lactulose exert an additive antitumorigenic effect in rat colon.