The using of four yeast strains (two cultures belonging to each S. cerevisiae and G. candidum) to withstand environmental conditions similar to the human digestion tract (probiotic criteria) were examined. The survival of these strains at low pH, in the presence of bile salt, in intestinal juice, also their ability to assimilate cholesterol were followed. However, the obtained results indicate that S. cerevisiae AAA3 and G. candidum GG1, may be promising candidate strains for use as probiotics. Thus, the potential role of these probiotic yeast cultures on serum lipid of rats was adopted. Twenty-four male albino rats were randomly and equally divided into four groups, six rats each. These rats were acclimatized on basal diet for 7 days before starting the experiment. The first group was fed on basal diet (cont. I), the second group was offered basal diet plus pasteurized buffalo's milk (6.5% fat). The rest two groups were fed on basal diet plus buffalo's milk in addition to either S. cerevisiae or G. candidum. Blood samples were collected at the beginning of the experiment (after adaptation period) and at the end of experiment. According to the result of serum analysis, total cholesterol and LDL- cholesterol levels in rats fed on S. cerevisiae (group 3) were decreased by 23% and 31.59% respectively. While, the corresponding values for rats consumed G. candidum were relatively lower, actually 21.46% and 29.12% respectively. Moreover, serum triglycerides concentrations reduced by 4.79% and 5.30% for rats in group 3 and 4, respectively as compared with control 1. Furthermore, data obtained declared that rats fed on either S. cerevisiae or G. candidum reduced markedly the atherogenic indexes 1 and 2 by mean values of 11.52, 23.86% and 10.1, 21.49% respectively, as compared with control 1. Continuously, rats fed on both yeast strains showed significant antagonistic effect against either coliform or staphylococci, where the reduction in their populations, actually 10.57, 7.42% for group 3 and 8.30, 18.22% for group 4, respectively.