It is well known that vaccination against paratuberculosis significantly reduces the number of clinical cases in field conditions. Different experimental studies have suggested that vaccination does not prevent infection, but modifies and limits the development of lesions. Histopathological methods have been demonstrated to be useful tools to assess the efficiency of paratuberculosis vaccines both in short term experiments and in field conditions. We have used these methods to evaluate vaccine efficacy on a system in which genetic factors have been minimized by the use of twin lamb pairs. One lamb out of each one of eight pairs of twin lambs was vaccinated when it was 15 days old. Fifty days later, all lambs were orally infected with a pathogenic ovine strain of Mycobacterium avium subspecies paratuberculosis. Three pairs were killed at 150 days post infection and the remaining 5 pairs at 350 d.p.i. On the postmortem, samples were taken for histopathological processing from ileocecal valve, three different segments of ileum, 3 distal, 3 medial, and 3 cranial jejunal Peyer's patches with their adjacent intestinal tissue and from the corresponding mesenteric lymph nodes. Clear differences were observed between vaccinated and non-vaccinated groups at both sampling times after infection. Whereas vaccinated lambs have only regressive-type granulomata limited to intestinal organized lymphoid tissues (Peyer's patches and ileocecal valve), non-vaccinated ones showed spread of granulomatous lesions to other areas of the intestinal wall. In conclusion, this experiment shows that vaccination induces noticeable differences within each pair of experimental animals, and it confirms the efficiency of this tissue sampling protocol since granulomata were located in the organized intestinal lymphoid tissue of all animals.