An estimated quarter of the global population has been infected with Mycobacterium tuberculosis (Mtb), the causative agent of tuberculosis (TB). Despite ongoing eJorts, TB reduction trends are only slightly declining. Mathematical modelling that recognises the spectrum of TB disease provides valuable insights into its policy implications. This thesis aims to generate accurate estimates of the burden of viable Mtb infection and evaluate the impact of population-wide screening interventions. Firstly, I assessed the impact of immunoreactivity test reversion on the estimated annual risk of infection (ARI), a key metric in TB epidemiology that measures Mtb transmission risk. When accounting for reversion, the true ARI was determined to be 2–5 times higher than previously estimated. Secondly, I refined previous Mtb infection estimates using a mathematical model that incorporated reversion-adjusted ARI trends, age-specific mixing, and self-clearance of infection. This analysis estimated that 2%—equating to 156 million people—are recently infected with viable Mtb and at high risk of disease progression. Thirdly, I re-evaluated the eJectiveness of mass screening interventions using chest radiography (CXR) as a strategy to significantly reduce TB prevalence. Finally, I calibrated a model to TB epidemiology in Vietnam and designed various population-wide screening algorithms to evaluate their impacts and costs. While a CXR-only algorithm rapidly reduced TB prevalence, they incurred high costs due to overtreatment. A combined approach of CXR followed by confirmatory bacteriological testing proved cost-eJective and became cost-saving compared to a business-as-usual counterfactual after the intervention ended. Overall, the findings of this thesis quantify the population eligible for TB preventive therapy and oJer insights into cost-eJective strategies for significantly reducing TB prevalence through population-wide screening in high-burden countries.