One reason proposed for the failure of Mycobacterium bovis bacille Calmette Guérin (BCG) vaccination to adequately control the spread of tuberculosis is a limited ability of the vaccine to induce effective CD8 T cell responses. However, the relative capacity of the BCG vaccine and virulent Mycobacterium tuberculosis to induce activation of CD8 T cells, and the factors that govern the initial priming of these cells after mycobacterial infection, are poorly characterized. Using a TCR transgenic CD8 T cell transfer model, we demonstrate significant activation of Ag-specific CD8 T cells by BCG, but responses were delayed and of reduced magnitude compared with those following infection with M. tuberculosis. The degree of CD8 T cell activation was critically dependent on the level of antigenic stimulation, as modifying the infectious dose to achieve comparable numbers of BCG or M. tuberculosis in draining lymph nodes led to the same pattern of CD8 T cell responses to both strains. Factors specific to M. tuberculosis infection did not influence the priming of CD8 T cells, as codelivery of M. tuberculosis with BCG did not alter the magnitude of BCG-induced T cell activation. Following transfer to RAG-1 -/- recipients, BCG and M. tuberculosis-induced CD8 T cells conferred equivalent levels of protection against M. tuberculosis infection. These findings demonstrate that BCG is able to prime functional CD8 T cells, and suggest that effective delivery of Ag to sites of T cell activation by vaccines may be a key requirement for optimal CD8 T cell responses to control mycobacterial infection.