Current control strategies are failing to contain the tuberculosis (TB) epidemic and are limited by our lack of understanding of Mycobacterium tuberculosis (M.tb) transmission dynamics, especially in high HIV prevalence settings. M.tb infection in children aged under 5 years of age indicates recent transmission, acting as a sentinel for infectious (typically adult) TB and highlights recent failures in community control measures. The overall aim of this research, which was based in a high HIV prevalence rural community in northern Malawi, was to delineate M.tb transmission events occurring within the context of a wellimplemented TB control programme, and thereby elucidate factors driving transmission that are not being addressed by current control strategies. This thesis presents findings from a series of linked studies, including a longitudinal tuberculin skin-test (TST) study of pre-school children in an area under demographic surveillance, and a household contact study of smear-positive TB cases. Estimates of the average annual risk of M.tb infection (ARTI) in this population of young BCG-vaccinated children varied widely depending on the method used to estimate infection prevalence. A previously overlooked method of estimating M.tb infection prevalence, initially published by Rust and Thomas in 1975, appeared to be only method to appropriately adjust for the marked effect of BCGattributable induration in the youngest children (aged <2 years). Marked differences in the estimates of the risk of M.tb infection when using crosssectional data compared to using longitudinal data are also highlighted. Age, known contact with a smear-positive TB case and community M.tb exposure (defined as the average notification rate of smear-positive TB per 100,000 population) were risk factors for prevalent and incident M.tb infection in children. Being HIV-exposed in utero was the strongest risk factor for prevalent infection, whilst having an HIV-positive father was strongly associated with incident infection. Additional risk factors for incident infection included church attendance and travel on mini-buses. No evidence was found in the household contact study that smear-positive tuberculosis patients on antiretroviral treatment (ART) were more likely to transmit M.tb infection to household child contacts (as inferred from TST positivity) compared to smear-positive tuberculosis patients not on ART. However, child contacts of HIV-negative individuals had nearly three times the odds of having a positive TST compared to child contacts of HIV-positive TB patients not on ART; this was partly explained by differences in the degree of smear positivity. HIV-related risk factors for prevalent and incident M.tb infection in pre-school children highlight that interventions, such as screening for HIV and TB with implementation of isoniazid preventive therapy where indicated within the household of HIV-positive individuals may further reduce the burden of TB at a community level in the longer term. Findings from this research also highlight the need for better infection control practices (improved ventilation) in congregate settings such as churches and mini-buses. There is an urgent need for improved ongoing surveillance to guide the implementation of context-specific TB control strategies.