Background: Tuberculosis is a leading cause of death worldwide from a single infectious agent, with an estimated 10.6 million new cases and 1.6 million deaths in 2021. The global burden of tuberculosis is highest in low- and middle-income countries (LMICs), and the largest burden of disease by country is found in India, accounting for around 30% of cases and deaths in 2021. New tuberculosis vaccines could play a key role in reducing the global burden of tuberculosis. As of 2023, there were sixteen candidates in clinical trials, and the vaccine candidate M72/AS01E and BCG-revaccination are of particular interest, as they have recently completed Phase IIb trials with positive results. There is optimism that a new vaccine or a recommendation for policy change for an existing vaccine could come within the next decade. It is important for decision makers to have information about the potential health and economic impacts of tuberculosis vaccines to support development and planning for vaccine introduction. However, decision makers will require different information to make informed decisions depending on their role. Global decision makers need to know the potential value of introducing tuberculosis vaccines to encourage continued development of candidates, whereas decision makers for national and subnational governments need information on country-level implementation of vaccines, such as M72/AS01E and BCG-revaccination, and the impact of any regional specific differences. The overall aim of this thesis is to use mathematical modelling to generate appropriate evidence to provide decision makers globally, and at various levels of government in India, with estimates of health and economic impact to support tuberculosis vaccine development, policy, and introduction. Methods: I developed a new tuberculosis model to incorporate recent advances in the knowledge of tuberculosis natural history, including subclinical disease and self-clearance. To support decision makers globally and encourage investment in vaccine candidates, I independently calibrated the model to 105 LMICs representing 93% of the global tuberculosis burden. I estimated the potential health impacts of vaccines meeting the WHO Preferred Product Characteristics for New Tuberculosis Vaccines by WHO region, World Bank income group, and WHO burden group. To support country-specific decisions regarding vaccine delivery of M72/AS01E and BCG-revaccination in India overall, I developed a sophisticated country-level model to simulate tuberculosis vaccine introduction in India. The country-level model incorporated differences in public and private sector treatment, and calibration targets over time to constrain long-term dynamics. I simulated M72/AS01E and BCG-revaccination scenarios exploring uncertainty in product characteristics and delivery. I estimated reductions in tuberculosis cases and deaths by each scenario, as well as costs and cost-effectiveness from the health system and societal perspectives. Finally, to assess potential differences between delivery strategies for M72/AS01E and BCGrevaccination within India, I developed subnational models for Delhi and Gujarat—two regions within India estimated to have the highest and lowest adult prevalence of tuberculosis disease respectively, incorporating regional specific differences in demography, infection prevalence, and treatment. I simulated M72/AS01E and BCG-revaccination scenarios to quantify and compare the predicted health and economic impacts. Results: Results from the multi-country modelling suggest that introducing new tuberculosis vaccines aligned with the WHO Preferred Product Characteristics could have a substantial impact in LMICs. Rapid introduction of an adolescent/adult vaccine could prevent up to 76.0 million cases and 8.5 million deaths before 2050. A booster infant vaccine could prevent up to 18.0 million cases and 2.6 million deaths before 2050. Evaluating the delivery of specific vaccine candidates through country-level modelling in India, M72/AS01E vaccines could avert up to 19.3 million cases and 3.1 million deaths, and BCGrevaccination could avert up to 15.2 million cases and 2.6 million deaths by 2050. Nearly all vaccine scenarios were cost-effective at the most conservative country-level threshold compared to no new vaccine introduction, and the average annual cost of M72/AS01E vaccination was around four times greater than that of BCG-revaccination. When comparing delivery strategies for specific vaccine candidates using the subnational models within India, M72/AS01E vaccines could avert up to 1.0 million cases in Delhi and 575 thousand cases in Gujarat, and BCG-revaccination could avert up to 626 thousand cases in Delhi and 222 thousand cases in Gujarat. Both subnational models predicted a larger impact of M72/AS01E vaccines than the impact of BCG-revaccination. The relative impact of BCG-revaccination for scenarios delivering the vaccine to older ages or all adults was higher in Gujarat compared to Delhi. Conclusions: New tuberculosis vaccines are likely to have a substantial health and economic impact globally and in India. The considerable reduction in cases and deaths estimated from multi-country modelling of tuberculosis vaccines in 105 LMICs supports the case for sustained investment in tuberculosis vaccine candidates throughout the pipeline to ensure success. While uncertainty in the actual vaccine characteristics exists, the modelling suggests that including a vaccination campaign, as well as rapidly introducing and scaling-up a new vaccine is crucial to obtain rapid impact by 2050. Country-level modelling in India and subnational modelling in Delhi and Gujarat of M72/AS01E and BCG-revaccination were estimated to be impactful and cost-effective. Knowledge of the population demographics and age-specific prevalence of infection will be important for decision makers in India to consider when considering delivery strategies for vaccines which are not effective for all infection statuses. Overall, these results suggest tuberculosis vaccines could be impactful and effective no matter the geographic scale of analysis, but at the national and subnational level, vaccine delivery strategies need to take into account regional variation in epidemiology and health care access to allow for the greatest possible impact. The evidence generated in this thesis has and can support global decision makers with vaccine investment decisions, and the Indian government with decisions regarding policy and vaccine delivery.