Paratuberculosis (PTBC) represents a burden for productivity in some sheep flocks in Spain. We thought that studying the methods of diagnosis and improving our knowledge on the epidemiology of the disease in this species could help us to develop suitable measures of control. Six flocks with a history of clinical cases of PTBC were included in a follow-up program, in which several tests (immunodifusion [ID], ELISA and fecal culture) were carried out at least one per year. The size of the flocks ranged between 100 and 250 ewes, making a total of 736 animals at the beginning of the study. Up to now seven clinical cases have been reported, representing about 0.64% of the mean number of the nonvaccinated animals. Serlogical tests have been performed twice. The ID detected 3.0% the first year and 2.3% the second, for an average of 2.7%. The ELISA gave positive results in 7.9% of the animals in the first test and 28.3% in the second. Taking into account all the bands which appeared in the ID, both methods showed a significant correlation. Fecal culture, performed in 167 samples, only yielded one positive result (30 weeks of incubation). The survival rates at the second test for the ID and ELISA positive ewes, were 50.0% and 66.7% , respectively, against 66.6% and 66.1% for the negative ones. Only 50.0% of the positives in the first ID test were still positive in the second, while 77.3% of the ELISA postives in the first test remained positives in the second. Fecal culture detected less positives than ID test, and did not detect the only clinical case for which culture was performed about a year before. The ELISA, showed a sharp increase in postivity in the second year, that could not be related with any factor. Thus, with some drawbacks, the most practical test appeared to be the ID, which detected clinical cases between twelve and three months before starting symptoms.