Current diagnostic tests for tuberculosis (TB) are not able to distinguish active disease from latent Mycobacterium tuberculosis infection, nor are they able to quantify the risk of a latently infected person progressing to active TB. There is interest, however, in adapting antigen-specific gamma interferon (IFN-gamma) release assays (IGRAs) to predict disease outcome. In this study, we used the differential susceptibilities of inbred mouse strains to M. tuberculosis infection to evaluate the prognostic capabilities of IGRAs. Using lung and blood cultures, we determined that CBA/J, DBA/2, and C3H/HeJ mice (models of heightened risk of progression to active TB) produced less antigen-specific IFN-gamma in response to M. tuberculosis culture filtrate proteins and early secreted antigenic target-6 than the relatively resistant C57BL/6 mouse strain. Additionally, reduced IFN-gamma secretion in supernatants reflected a reduced frequency of IFN-gamma-responding cells in the lung and blood and not a specific defect in IFN-gamma secretion at the single-cell level. Importantly, detection of antigen-specific IFN-gamma from blood cultures accurately reflected lung responses, indicating that blood can be an appropriate test tissue in humans. Furthermore, reduced antigen-specific IFN-gamma production and low frequencies of IFN-gamma-responding cells from peripheral blood predicted increased risk of TB disease progression across genetically diverse TB disease-susceptible mouse strains, suggesting that similar results may occur in humans. The development of efficacious predictive diagnostic tests for humans would lead to targeted therapy prior to progression to active TB, reducing transmission, incidence, and prevalence rates while maximizing the use of public health resources.