Around 11 million people developed tuberculosis (TB) in 2021, and 1.6 million died from the disease. Current control strategies are insufficient, with global TB incidence falling by only 2% per year. One reason for the slow decline may be widespread reliance on passive case detection - requiring people with TB to present to healthcare services with symptoms. This means that people can be infectious for months or years before diagnosis, and an estimated 40% of incident TB was not diagnosed in 2021. Active case finding (ACF) - the systematic screening of high-risk groups or populations - is one way to find people with TB earlier, leading to reductions in transmission. The World Health Organization recommends ACF in areas with a high prevalence of TB. Recent National Strategic Plans from countries as diverse as South Africa, Uganda, and India contain plans to scale-up ACF in high risk populations. Despite the scaling up of ACF activities, considerable uncertainty remains as to their likely impact, and how it varies between approaches and settings. Three randomised control trials (RCTs) estimating the impact of ACF on transmission have been conducted. One trial achieved an impressive 50% (95% CI 22-68%) reduction in the prevalence of infection in children (a proxy for transmission), demonstrating that community ACF can be a highly effective in reducing transmission. The other trials used less intensive intervention approaches, and found no evidence for reductions in transmission. A fourth RCT found a reduction in TB prevalence, but did not estimate reductions in transmission. Mathematical modelling suggests that the differences between the trial results cannot be explained by differences in the tests used or numbers of cases detected. There is a need to understand factors that affect the reductions in TB incidence achieved through ACF, and to identify less intensive and expensive ACF approaches that can lead to reductions in transmission. Mathematical modelling can be used to predict the impact of ACF on TB incidence. However, assumptions typically made in models may not be correct, and models of ACF have rarely been validated using empirical data. In particular, we have identified three factors that may alter the impact of ACF on TB incidence: A) People who have been screened in previous rounds may be more or less likely to seek or accept screening. B) Coverage tends to be lower in men than in women, despite higher TB prevalences in men. C) The probability of participating in ACF may be higher for people who were closer to seeking care and receiving a diagnosis passively. The impact of these factors may vary by intervention design and setting.