Saccharomyces boulardii (Sb), a nonpathogenic yeast, has been used to prevent recurrences of Clostridium difficile (C.diff) -associated diarrhea. A single report suggested that treatment with Saccharomyces cerevisiae (Sc), commonly called brewer's yeast (BY), facilitates treatment of persistent C.diff infection. We conducted this experiment to determine whether C.diff toxin A-induced colonic secretion in the rat is blunted by pretreatment with either Sb or BY. We employed closed cecal pouches in two groups of five adult male Sprague-Dawley rats fed with standard chow for five days prior to the experiment, another group whose water was supplemented with 20 x 10(9) colony-forming units (CFU) of Sb per day for five days, and another group whose water was supplemented with 20 x 10(9) CFU of Sc per day for five days. Cecal pouches were infused for 3 hr with one of the following: (1) normal saline alone for a control group, or (2) normal saline plus 5 microg of C.diff toxin A (for the other control group and for the two experimental groups). Water movement was measured by a nonabsorbable marker technique. Sodium movement and permeability to mannitol were also measured. Prior to the infusion, cecal contents were quantitatively cultured. In the three animals whose ceca were colonized with less than 10(6) CFU of either yeast per gram wet cecal content, toxin A-induced secretion could not be attenuated. In contrast, animals whose ceca were colonized with more than 10(6) CFU of either yeast per gram of wet cecal content showed significantly less secretion after toxin A application than those which were not fed yeast. S. cerevisiae reduced secretion by half (N = 5, P = 0.039 for water, 0.044 for sodium) and Sb by 75% (N = 4, P = 0.015 for water, 0.034 for sodium). Toxin-induced increases in permeability to [3H]mannitol from systemic circulation to cecum could not be blunted by either yeast. We conclude that rat ceca can be colonized by either organism and that both organisms reduce C.diff toxin A-mediated secretion. We speculate that both organisms might have benefit in human C.diff-associated enterocolitis. Further studies of their mechanisms of action as well as clinical trials for the prevention and treatment of human C.diff infections should be pursued.