Neglected Tropical Diseases (NTDs) are both causes and manifestations of poverty in developing countries. Recent advocacy efforts have increased the profile of NTDs, and led to bold new control and elimination targets set for 2020 by the World Health Organisation. However there are multifaceted challenges in effectively implementing NTD interventions in resource-poor contexts that need to be understood and engaged. While there is a growing call by researchers and international agencies for a science of global health delivery to understand these complexities, the exact nature of this science remains contested. This thesis contributes to these debates by advancing a critical social science perspective on the factors that mediate intervention effectiveness for NTD control. Grounded in a social constructivist approach using mixed methods, it critiques prevailing orthodoxies by unpacking the nature, processes and outcomes of three large-scale NTD prevention programmes in Eastern Africa. Focused on different diseases, these case studies represent different types of intervention approaches: top-down, participatory and public-private partnership. The thesis traces the social, technical and environmental processes that mediate the delivery, adoption and use of particular health technologies, such as pit latrines, insecticides and vaccination. Together, these case studies reveal surprisingly similar reasons for why many interventions do not perform according to expectations. Despite new approaches that claim to overcome stereotypical challenges of top-down planning, narrow technocratic perspectives continue to play a defining role in maintaining disjunctions between global aspirations, local realities and intervention outcomes. New perspectives and changes in orientation are needed that emphasise flexibility, learning and adaptability to local contexts. Towards this end, the thesis outlines a conceptual framework based on a comparative analysis of the case studies that highlights five interrelated domains where effectiveness is determined: geographical/livelihood variation, local agency, incentives, the socio-materiality of technology and planning/governance. I argue that addressing the shortcomings of contemporary interventions requires that programme planners actively engage these domains by seeking to “order complexity.” Greater integration of social science perspectives into the management of NTD programmes would provide significant benefit. In these ways, the thesis contributes to wider debates about the nature of global health interventions and the influence of local contexts in mediating efforts to improve the health and wellbeing of the world’s poor and marginalised.