The infectious nature of tuberculosis (TB) makes early TB detection an important public health objective. In the South African context, where the high TB burden remains concentrated among the poor and is exacerbated by the human immunodeficiency virus epidemic, TB control is a high priority for the National Department of Health. However, without an adequate grasp of the supply and demand issues underlying delays in TB detection, policies intended to improve detection, and ultimately TB outcomes, are unlikely to succeed. The overarching aim of the three chapters in this thesis is to consider the role of crucial demand-and supply-side constraints on early TB detection in South Africa. In this regard, three key factors are considered: The health-seeking behaviour of TB-symptomatic persons, the quality of TB care at the primary healthcare level and the role of nurse workloads in TB-detection protocol compliance. Early TB detection advances access to treatment, thereby preventing further transmission, in other words, demand prevention. Of great importance in determining whether a TB-symptomatic person seeks healthcare or not are the individual’s social and economic circumstances, as well as other generally unobservable factors such as TB stigma. Findings from Chapter 2 indicate that most persons with a chronic cough did not seek care for it. This is a perturbing finding considering that data were collected in high TB burden communities. Findings also provide evidence that in these communities, TB-control interventions may benefit from focusing on youth, those with lower levels of education, smokers and higher socioeconomic status sub-groups. Although no role for stigma in health-seeking behaviour was found, this finding underscores the need to improve methods to measure stigma. On the supply side, access to affordable, quality healthcare services with adequate capacity and the willingness to identify presumptive TB patients are essential prerequisites for the timely detection of TB. Chapters 3 and 4 explore supply-side weaknesses that place critical constraints on the effectiveness of TB detection at the primary healthcare level. Chapter 3 uses the unannounced Standardised Patient (SP) method to measure the quality of TB screening at primary healthcare facilities in urban South Africa. Findings from this chapter highlight the disconnect between the prescribed TB protocols and its implementation. Even though gaps and missed opportunities for early TB detection are identified (i.e. there is room for improvement), some findings are positive. These positive findings imply that there is a stronger than expected responsiveness to TB detection in South Africa than the current literature would suggest, although it is clear that weaknesses remain. Chapter 4 uses an SP fixed effects model to estimate primary healthcare facility correlates of TB-detection protocol compliance. The chapter provides evidence that human resource constraints at primary healthcare facilities are associated with suboptimal quality of TB detection. This chapter also recommends further investigation regarding the role of management in quality of care. The findings of this thesis contribute to the small but growing literature on the economics of infectious diseases and are intended to inform, guide and further enhance TB policies in South Africa. It also provides quantitative evidence for future quality-improvement research in this area.