Several clinical cases of paratuberculosis in dairy cattle and sheep were recently confirmed in our laboratory, nevertheless the accurate prevalence data for northern Mexico is lacking. A total of 303 individual blood samples from 37 dairy cattle herds in Tijuana, Baja Claifornia, Mexico were analyzed to estimate the seroprevalence of bovine paratuberculosis. These samples are representatives of approximately 20,000 cows belonging to the local Dairy Producers Association in Tijuana, Baja California (AGLPLT). They were taken as a part of a seroprevalence survey of major infectious abortive diseases during the period from December 2006 to March 2007. Herd size ranged approximately from 100 to 1000. Collected sera were subjected to in-house ELISA test by using commercial PPA antigen and control reagents (Allied Monitor Inc., USA). Ninety-one samples were also tested by a commercial kit (Institut Pourquier, France). For the in-house ELISA, 28 sera gave positive results, 27 were suspicious, and 248 were negative. Apparent prevalence of infection at the animal level was 9.2% (95% I.C. 6.5-13.0%). Using a sensitivity value of 58.8% and a specificity of 95.4% (as claimed by the manufacturer) the overall true prevalence in dairy cattle in Tijuana was estimated as 8.6% (95% I.C. 3.2-15.9%). Apparent herd-level prevalence was 43.2% (16/37, 95% I.C. 28.7-59.1%) when the herds with at least 1 positive animal were considered as positive. When the herds with only 2 or more seropositive animals were considered positive, the prevalence was shifted to 21.6% (8/37, 95% I.C. 11.4-37.2%). These results are consistent with the precise information of the disease infection in California (Adaska and Anderson, 2003). Since the importation of the cattle from US to Mexico is currently limited and aggressive culling is not employed in the local herds, the immediate implementation of disease control program is necessary to prevent the economic loss due to the disease propagation in this region.