Tuberculosis (TB) chemotherapy eliminates disease symptoms, but often fails to sterilize Mycobacterium tuberculosis (Mtb) leaving apparently cured TB patients at risk of disease relapse. How Mtb survives during post- treatment persistent infections (PTPI) is ill defined and we lack a full understanding of the immune determinants that prevent progression to active TB (relapse). We developed a paucibacillary (low bacterial numbers) mouse model that mimics PTPI in in humans. We found that silencing of in vivo essential genes cured Mtb infections in mice to the extent that colony forming units (CFU) were no longer detected on agar plates. However, as in humans, mice were often not sterilized, and paucibacillary infection resulted in relapse TB. Our model allows us to interrogate both, the mechanisms that enable Mtb to persist despite host immunity and the host factors that are required to control and sterilize Mtb infection. We will use this model to (1) compare post-treatment persistent infection (PTPI) in mice and humans (2) determine the importance of specific leukocyte subsets for the establishment or control of paucibacillary Mtb infection and (3) identify Mtb genes that are required for paucibacillary persistence.