Although Lyme disease is currently classified as exotic in Mexico, recent studies have suggested that it might be endemic there. We assessed the potential risk for the establishment of Borrelia burgdorferi transmission in Mexico. To identify the potential routes of B. burgdorferi spread, Complex Inference Networks were used initially to identify potential vector-host interactions between hard ticks (Ixodes) and migratory birds in the U.S., and a model for predicting the most important potential bird hosts of hard ticks was then obtained. By using network metrics, keystone-vectors were identified as those species with highest connectivity within and between network communities and had the potential to keep the pathogen circulating with many birds and to be dispersed to several regions. The climatic profile where these interactions occur in the U.S. was characterized and a geographic model for each keystone-vector was built. The accuracy of these models to predict areas where hard ticks have been reported positive for B. burgdorferi allows one to identify areas of greater risk of Lyme disease emergence. These hard tick-bird interactions and their climatic profile were mapped into the winter ranges of birds in Mexico. Thus, those regions in Mexico with the highest potential for becoming endemic areas of Lyme disease through the arrival of hard ticks and birds infected by B. burgdorferi were identified. These areas are candidates for future surveillance programs.