The South African tuberculosis (TB) epidemic is driven mainly by HIV, and the TB disease burden is greater in males than females. Additional factors that drive the epidemic include undiagnosed and untreated TB, contributing to transmission; and highly prevalent TB risk factors such as alcohol misuse, smoking, diabetes, and undernutrition, which increase the risk of progression to TB disease. These factors are distributed differently by sex and likely explain the observed sex disparities in TB. The South African TB control programme has implemented multiple interventions, including directly observed therapy strategy (DOTS), antiretroviral therapy (ART), intensified screening activities, the provision of isoniazid preventative therapy (IPT) and the implementation of Xpert MTB/RIF as a first-line diagnostic tool. However, few analyses have quantified the historical impact of HIV and the combined impact of TB interventions on the South African TB epidemic at a national level. In addition, factors that influence sex disparities in the South African TB burden have not been explored thoroughly. Also, it remains uncertain whether, with existing interventions, it would be feasible for South Africa to meet the End TB targets to reduce TB incidence and mortality by 80% and 90% respectively (relative to 2015 levels) by 2030. This thesis aims to address the abovementioned gaps in knowledge and provide insights into understanding the population-level TB dynamics, using a mathematical model. The first objective is to quantify TB incidence and mortality due to HIV and assess the impact of interventions mentioned above on TB incidence and mortality between 1990 and 2019. The second objective is to explore the extent to which the following factors contribute to sex differences in TB: HIV, ART uptake, smoking, alcohol abuse, undernutrition, diabetes, health-seeking patterns, social contact rates and TB treatment discontinuation. The third objective is to project the future impact of increasing screening, improving linkage to TB care and retention, increasing preventative therapy, and reducing ART interruptions. An age- and sex-stratified dynamic tuberculosis transmission model for South Africa was developed. To dynamically model the effect of HIV and ART on TB incidence and mortality, the TB model was integrated into the Thembisa model, a previously developed HIV and demographic model. In addition, age- and sex-specific relative risks were applied to rates of progression to TB disease to capture age and sex differences in tuberculosis incidence. The model also included a diagnostic pathway representing health-seeking patterns and the sensitivity and specificity of the diagnostic algorithm. A Bayesian approach was used to calibrate the model to the numbers of people starting treatment from the electronic tuberculosis register, deaths from the vital register, microbiological tests, and the national tuberculosis prevalence survey. The model estimated rapid increases in TB incidence and mortality in the mid-to-late 1990s, influenced by HIV. Between 1990 and 2019, approximately eight million people developed tuberculosis, and two million died from TB; HIV accounted for at least half and two-thirds of the TB incidence and mortality, respectively. The TB epidemic peaked in the mid-to-late 2000s, followed by declines until 2019. The ART program and TB screening efforts, which were expanded in the mid-2000s, contributed the most to reductions in TB incidence and mortality, while other interventions had minor impacts. Due to the heavier HIV burden in women than men, women experienced greater HIV-associated TB incidence and mortality than men. However, because of the higher ART uptake among women than men, women experienced greater relative reductions in TB incidence and mortality over the period 2005– 2019. Consequently, the higher TB burden among men has been sustained; the estimated male-to-female ratios of TB incidence and mortality in 2019 were 1.7 and 1.65, respectively. Additional factors explaining the excess TB in men are smoking, alcohol abuse and delays in health-seeking patterns. Sex differences in undernutrition, social contact patterns, and treatment discontinuation had minimal effect on TB sex disparities. Projections of the model to 2030, considering the effects of COVID-19-related disruptions to TB care, suggest that increasing TB screening would be the most impactful among all interventions explored. However, the model also suggests that the 2030 End TB milestone is unlikely to be met by scaling up existing interventions. Other interventions that need to be explored include targeted universal TB testing and other diagnostic tests such as digital chest x-rays, urine Lipoarabinomannan, and biomarkers to identify individuals at risk of TB disease. Accelerating progress toward TB incidence and mortality reductions will require developing affordable and efficient rapid diagnostic tools to identify potential and active TB cases. Research and innovation efforts towards finding a vaccine effective in preventing TB disease are also critical. In addition, it is essential to improve the uptake of TB preventative therapy in HIV-positive individuals and perhaps further expand provision to other TB risk groups