We addressed the question of whether protective immunity induced by natural infection with Mycobacterium tuberculosis and that induced by vaccination with Mycobacterium bovis bacille Calmette-Guerin (BCG) differ in the murine model. We infected mice with M. tuberculosis Erdman, cured them by chemotherapy, and subsequently reinfected them with a low dose of M. tuberculosis H37Rv. The course of tuberculosis was compared with that in mice previously vaccinated with BCG Danish 1331. Protection against postprimary M. tuberculosis infection did not differ significantly between the 2 groups. After challenge infection, numbers of interferon- gamma -positive splenocytes did not differ between mice with primary infection and vaccinated mice. Splenocytes from primary M. tuberculosis-infected mice conferred marginally higher protection than did those from BCG-vaccinated mice. Serum transfer did not protect against reinfection in either group. Our data emphasize that natural infection with M. tuberculosis and vaccination with BCG do not differ in their capacity to induce protective immunity against tuberculosis and support the notions that reinfection contributes to the development of active disease and that any novel vaccine against tuberculosis has to perform better than both vaccination with BCG and immunity evoked by natural infection.