Aerosols containing Mycobacterium tuberculosis (MTB) generated from the cough of patients with active pulmonary tuberculosis are the source of MTB infection. About 70% of individuals exposed to infected aerosols do not get infected, depending on the intensity and duration of MTB exposure. Only 40% of the rest of the individuals (about 10% of those originally exposed) develop primary tuberculosis, whereas the remaining 60% contain the infection with generation of a robust immune response leading to latent tuberculosis, which is regarded as a spectrum rather than a single entity. The mechanisms involved in this natural protection are not yet well understood. There is an increasing need to integrate all disparate observations into a coherent systems biology approach for a comprehensive understanding: we need to decipher the nature of success and failure in MTB infection in humans. New advances in cellular immunology will aid in achieving that goal. We review here the nature of MTB peptide generation, antigen presentation, and detection of major histocompatibility complex class I and II-presented T-cell epitopes. Cross-sectional thinking from lessons learned in the context of the major efforts to develop vaccines will help to dissect biologically relevant mechanisms that need to be translated into the clinical context of MTB infection with the aim to (1) better understand clinically relevant T-cell responses in individuals protected from tuberculosis disease and develop markers of immune protection and vaccine take, (2) characterize the nature of the immune response in individuals who are not able to contain MTB infection, and ultimately (3) characterize markers to gauge response to therapy.