The control of vampire bat rabies (VBR) in Brazil is based on the culling of Desmodus rotundus and the surveillance of outbreaks caused by D. rotundus in cattle and humans in addition to vaccination of susceptible livestock. The detection of anti-rabies antibodies in vampire bats indicates exposure to the rabies virus, and several studies have reported an increase of these antibodies following experimental infection. However, the dynamics of anti-rabies antibodies in natural populations of D. rotundus remains poorly understood. In this study, we took advantage of recent outbreaks of VBR among livestock in the Sao Paulo region of Brazil to test whether seroprevalence in D. rotundus reflects the incidence of rabies in nearby livestock populations. Sixty-four D. rotundus were captured during and after outbreaks from roost located in municipalities belonging to three regions with different incidences of rabies in herbivores. Sixteen seropositive bats were then kept in captivity for up to 120 days, and their antibodies and virus levels were quantified at different time points using the rapid fluorescent focus inhibition test (RFFIT). Antibody titers were associated with the occurrence of ongoing outbreak, with a higher proportion of bats showing titer >0.5 IU/ml in the region with a recent outbreak. However, low titers were still detected in bats from regions reporting the last outbreak of rabies at least 3 years prior to sampling. This study suggests that serological surveillance of rabies in vampire bats can be used as a tool to evaluate risk of outbreaks in at risk populations of cattle and human.