New post-exposure tuberculosis vaccination strategies are being developed to prevent disease in individuals latently infected with Mycobacterium tuberculosis. However, concerns about the potential induction of deleterious Koch-like reactions after immunization of persons with latent tuberculosis has limited progress in assessing the effectiveness of post-exposure vaccination. To evaluate the safety of immunization after M. tuberculosis infection, two mouse models were established, a drug treatment low bacterial burden model and an active disease model. Twelve different M. tuberculosis antigen preparations and vaccines (including DNA, subunit, viral vectored, and live, attenuated vaccines) were evaluated using these mouse models. In the low bacterial burden model, post-exposure vaccination did not induce significant reactivational disease and only injection of BCG evoked increases in lung inflammatory responses at 1 month after the immunizations. Additionally, although significant increases in lung inflammation were seen for animals injected with the hps65 DNA vaccine or a M. tuberculosis culture supernatant preparation, no differences in the survival periods were detected between vaccinated and non-vaccinated mice at 10 months post-immunization using the low bacterial burden model. For the active disease model, significantly more lung inflammation was observed at 1 month after administration of the hsp65 DNA vaccine but none of the antigen preparations tested increased the lung bacterial burdens at this early time point. Furthermore, vaccination of diseased mice with BCG or TB DNA vaccines did not significantly affect mortality rates compared to non-vaccinated controls at 10 months post-immunization. Overall, these data suggest that while the potential risk of inducing Koch-like reactions is low after immunization of persons with latent tuberculosis, extreme caution is still needed as post-exposure vaccines progress from pre-clinical experiments into the initial phases of clinical testing.