Our hypothesis was that administration of bifidobacteria, Lactobacillus acidophilus or both to rats will minimize the numbers of aberrant crypts in the distal colon that develop in response to the carcinogen 1,2-dimethylhydrazine (DMH). A series of experiments was designed to test this hypothesis where the treatments used were as follows: skim milk controls (Skim-Basal), skim milk + bifidobacteria (Bifido-Basal), skim milk + fructooligosaccharide (Skim-FOS), and skim milk + bifidobacteria + fructooligosaccharide (Bifido-FOS). In two experiments, bifido-bacteria + FOS administration significantly decreased the number of aberrant crypts that developed, but there was no clear relationship of aberrant crypts to numbers of bifidobacteria or Clostridium perfringens. In the third experiment, the Bifido-FOS treatment led to significantly fewer aberrant crypts and aberrant crypt foci than the Bifido-Basal treatment. The Skim-FOS group had significantly more cecal bifidobacteria than the Skim-Basal group and significantly fewer C. perfringens than the Skim-Basal and Bifido-Basal. In a fourth experiment, L. acidophilus was added as an additional treatment. The number of aberrant crypts was not significantly different among the groups. However, the number of C. perfringens was significantly decreased by the addition of bifidobacteria, L. acidophilus or the combination of the two, whereas the numbers of bifidobacteria and L. acidophilus were not affected by treatment. A significant correlation (R2 = 0.84, P < 0.01) was noted between the body weight of rats at DMH administration and the magnitude of the difference in aberrant crypts between the Skim-Basal rats and the Bifido-FOS rats. The results suggest that there is variability in the effects of bifidobacteria and L. acidophilus administration on both aberrant crypt formation and C. perfringens.