Since the 1970s, female-headed households (FHHs) in developing countries have often been used in development literature as a proxy for poverty and vulnerability. In reality the profile of women-headed households is diverse; they include, at the least, rich and poor women, aged widows as well as young single mothers and wives of migrant workers, educated professionals and semi-literate manual labourers. This diversity of characteristics, with its attendant diversity of experience and vulnerability unfolds a picture of heterogeneity, rather than homogeneity. Yet, despite ample evidence that FHHs are, in fact, heterogeneous and not homogeneous, contemporary research and practice remains caught dominantly within the ‘poverty-vulnerability’ nexus. The heterogeneity of female headship is undermined by conventional notions of homogeneity. It is this gap that the present research addresses. Drawing on interdisciplinary perspectives especially from demography, gender studies; particularly gender and development, risk and vulnerability studies, as well as scholarship on social capital, this thesis employs a ‘heterogeneity’ lens to specifically examine the complexities of household formation, economic conditions and social relations of FHHs in Sri Lanka, in an attempt to explore their vulnerabilities and resilience. The choice of Sri Lanka as the context for this study is grounded in the demographic reality of a relatively high, and consistently increasing, proportion of households headed by women since the 1970s. By 2009/10, FHHs accounted for nearly one-quarter of all households, throughout the country. In order to capture the geographical and social diversity of FHHs, empirical research was conducted in three contrasting types of district in Sri Lanka, encompassing urban, rural, and estate sectors. Two main data collection strategies were employed in a mixed methods approach: a sample survey of a cross-section of 534 FHHs, and in-depth interviews with 32 female heads purposively selected from among the survey participants. The findings and discussions include quantitative statistical and qualitative thematic analyses based on primary data, combined with secondary data from censuses, national survey reports and micro-studies of FHHs in Sri Lanka. The key findings show the diversity in profile of FHHs in the sample: they range from single person to large extended households. While some households consist of only the woman head and her young children, others comprise aged parents and a woman head. Households were also constituted of ‘working-age’ household members, including the female heads who were totally reliant on others for income and other resources. The study also revealed novel findings that challenge the emphasis of most conventional perceptions of poverty and female headship. From an economic perspective, the results show women from rich households can be personally poor, lacking, among others, in skills to manage household economies, while women in low-income brackets may be resilient, enterprising and satisfied with their needs, despite their apparent poverty. Finally, the thesis highlights the significant role of social capital, a relatively under-researched area in relation to FHHs. The findings reveal that many female heads in Sri Lanka are rich in social capital, a resource in its own right for these women. However, social capital itself needs to be disaggregated into ‘support networks’ and ‘leverage networks’ to understand the role it plays in providing long-term security and resilience. The results show that the majority of FHHs in the sample had access to support networks that provide day-to-day subsistence, but which did not offer them prospects to leverage out of their current situation.