Mycobacterium tuberculosis causes more deaths in humans than any other bacterial pathogen. The most recent data from the World Health Organization reveal that over 9million new cases of tuberculosis occur each year and that the incidence appears to be increasing with population growth. Despite the global burden of tuberculosis, we are still reliant on relatively dated measures to prevent, diagnose, and treat the disease. New, more effective tools are needed to diminish the incidence of tuberculosis. M. tuberculosis lacks a natural host beyond humans and, hence, surrogate models have been employed in the study of the pathogen. The discovery and development of new vaccines, diagnostics, or antitubercular drugs are dependent upon the validity of any experimental model used and its relevance to tuberculosis in humans. In this review, a range of experimental models, from in vitro studies with fast-growing low-pathogenic species of mycobacteria to the infection of nonhuman primates with virulent M. tuberculosis, will be discussed.