Housing subsidies are a key strategy used by governments to bridge their low-income housing gap. Several low- and middle-income countries subsidise rural welfare housing – at times through direct cash transfer to recipients – conditional on achieving specific construction milestones through ‘self-help’. This is based on the presumption that recipients maximise benefits by choosing how to utilise the subsidy for house construction depending on their needs. Yet current approaches to ‘self-help’ rural welfare housing, with its technical focus on house-type specifications and quantitative focus on measuring success, fail to evaluate processes through which these subsidy schemes are implemented on-the-ground. Hence, this thesis examines implementation strategies from the perspective of three crucial ground-level actors–subsidy recipients (past, current and prospective), masons and field-staff–in the world’s largest self-help rural welfare housing scheme–India’s Pradhan Mantri Awas Yojana-Grameen, which targets building 30 million houses between 2016-2022 for people living under extreme poverty. Despite its commendable objectives and having delivered nearly 36 million houses (1996-2015), the scheme has reported persistent problems since inception (1996) in implementation and compliance with its spatial, temporal and disaster resilience requirements. To address these problems, whilst number of reforms have been introduced, the ‘self-help’ requirement–widely recognised as key for effective and efficient delivery–has persisted. Yet its actual ground-level practice and implications has remained underexplored and poorly understood in rural welfare housing scholarship. Using mixed methods through a constructivist grounded approach, this research examines the relationships between the Scheme’s requirements, its actual practice and its implications on recipients’ welfare in peri-urban and rural riparian settlements in Cuttack District in the coastal State of Odisha. Drawing on nearly a year of fieldwork, this study qualitatively analysed 77 semi-structured interviews with the three ground-level implementers, participatory card games with recipients, observations of construction methods with masons and field-staff, and detailed architectural documentation of house-types. Further, it quantified qualitative data and quantitatively analysed official secondary data sets for triangulation. The empirical findings reveal important aspects of the nature and welfare implications of trade-offs that ground-level implementers undertake for achieving compliance with the scheme requirements. Ground-level implementers were (a) overstretched in fulfilling on-the-ground highly gendered responsibilities within a context of information, time and resource scarcity; (b) caught in debt cycles partly due to untimely instalment releases, underestimated official construction costs, overestimated labour inputs and costs for ensuring smooth construction process; and (c) required to build amid severe space, material and construction tools constraints, leading to compensatory and non-standard construction practices, which influenced structural integrity and ultimately recipients’ aspiration and the Scheme’s objective of intergenerational resilience. These findings provide important directions for future interdisciplinary empirical research on social networks and trust, monitoring and change-agents, and financing and instalment structure, with implications for policy for promoting a culture of safety in Scheme house construction.