India's demographic transition has the potential to pressure its largely family- focused system of support for dependent older people, which may have negative implications for the availability of support for the older population. This thesis aimed to develop a nuanced understanding of the potential impact of India’s demographic transition for social support to (and subsequent health of) older people, considering variation across populations, and to recommend solutions to ensure support for India’s older generations, considering preferences of the population. To achieve these aims, I used a mixed-methods approach. I combined analyses of secondary quantitative data (2011 census data and National Sample Survey Organisation (NSSO) surveys (1995-96-2014)) with the collection and analysis of primary qualitative data in Tamil Nadu, a southern state that has relatively low fertility across socioeconomic strata. The sample consisted of a socioeconomically diverse group of N=113 adults (aged 20-64) with varying experiences of supporting older relatives. Chapter five described family sizes (number of children, sons, daughters) at the subnational level (by state, urban/rural residence and socioeconomic status) for ever married women aged 60-plus in 2011. This highlighted large state and socioeconomic differences in family sizes. In many states, women with the least education had smaller family sizes than those with some education, contrary to what might be expected from fertility trends. Chapter six described family structure trends (number of children, sons, daughters, and marital status) for older people (aged 60-plus) at the national level between 1995-96 and 2014 and examined the relationship between family structure and self-rated health. This indicated that, for the national average, family structures have not changed to a degree that might impact support (i.e., having zero children or sons remained rare in 2014) and that support was associated with positive health outcomes. Chapter seven used the primary qualitative data to understand how older people are supported in Tamil Nadu, which demonstrated some similarities in support between socioeconomic groups (for instance the responsibility tended to fall on the closest child-spousal unit) and some differences (e.g., the use of formal care, daughters' support). Chapter eight used the qualitative data to explore attitudes around varying support arrangements and preferences for own (future) support, which indicated that people highly valued the co-resident family-focused arrangement for the provision of both tangible support as well as demonstration of love and care. Nevertheless, they were consistently pessimistic about the availability of support from their children in the future. Finally, chapter nine used the qualitative data to understand the challenges that family members experience when supporting their older relatives and the ways in which they cope, which identified a range of stressors (some related to Tamil Nadu’s demographic transition) and differences in the coping strategies available to varying socioeconomic groups. Based on these combined findings, I have concluded that fertility decline will reduce the support available to dependent older individuals that lack the resources to adapt (i.e., those of lower socioeconomic status and/or rural) with negative implications for their health, as it (a) will increase the chance of being sonless (and to a lesser extent, childless) and (b) will reduce the pool of children who are both willing and able to support. The timing of these effects will vary greatly between regional and socioeconomic groups, many lower socioeconomic status older individuals already receive limited support due to socioeconomic pressures that restrict the support their children can provide them. Given the high value assigned to the family system of support, I have proposed that policy should (a) primarily aim to reduce the difficulties experienced by family carers for their own wellbeing as well as to promote the family-based support available to (and health of) older dependent individuals, as well as (b) provide financial and practical support for older individuals for whom family-based support is unappealing or unavailable in a culturally acceptable manner , and (c) improve people’s ability to remain financially and physically independent in their later years. I have suggested strategies for achieving these three aims, of which universal health coverage is a key component.