Cellular immunity plays an important role in the pathogenesis of paratuberculosis. To determine the changes in lymphocyte populations associated with vaccination, a sample of vaccinated and non-vaccinated (positive fecal smear and normal) sheep were submitted to a FACS analysis using a panel of 11 monoclonal antibodies (CD2, CD3, CD4, CD8, CD21, CD31, CD41/61, CD43, CD45R, and gamma-delta TCR). Results were transformed and submitted to statistical analysis by analysis of variance and principal components multivariate analysis. Significant differences between vaccinated and non-vaccinated sheep were found for CD45R and CD3 markers. Vaccinated sheep showed higher proportions of CD45R and lower of CD3. The same tendencies were observed for normal and for positive sheep when both groups were analyzed separately, although the latter usually showed slightly higher counts than the former. All the markers were positively correlated with the first component of the principal components analysis. However, the second component, although defining a weaker relationship, seemed to define two groups of markers: CD43, CD45R, and CD21 on one side and the rest on the other. When considering the clustering by sample group, the patterns of distribution varied considerably. The most relevant observation was the defined opposition between CD45R and CD2 along the first component in the normal group. In the positive group, a similar situation was observed, but between CD45R and CD3 instead of CD2. These relationships were confirmed by Pearson correlation analysis which showed that CD45R/CD3 correlation was negative and significant (r=0.996, p<0.06). In the positive group, these relationships became -0.022 (N.S.) and -0.999 (p<0.02), respectively. Analyzing the clustering of sheep according to their CD patterns showed a certain segregation of sheep on the second component. Since the low number of sheep made this separation little clear, another set of non-vaccinated sheep (4 positive and 3 normal) with incomplete data was added to the original data. In these conditions, the same pattern of segregation appeared more clearly, with 80% of vaccinated sheep on the left side of the second component, and 100% of normal on the right (p<0.001). Positive sheep split between both sides with 33% on the left and 67% on the right. These results show characteristic changes in the cellular immune response to M. a. paratuberculosis sensitization, with CD45R and CD3 appearing as the subpopulation more clearly involved in vaccine protection. The differences between normal and vaccinated CD patterns stresses the relevance of cellular immunity. However, the observation that shedding infected sheep do not specifically conform to any of the two patterns suggests complex cellular relationships in the pathogenesis of this infection.