Researchers often face problems when trying to unpack the causal pathways embedded in complex social processes, such as between poverty, housing, and poor health. Difficulties in pinning down cause and effect relationships and identifying how these relationships work in different contexts make it challenging to understand how one set of circumstances, action, or intervention could lead to a particular outcome. However, it is important to improve our understanding of these processes, so that we can influence the design of health and social care systems that work to improve health and reduce inequalities. In this case study, we reflect on our research into the health and wellbeing impacts of different approaches to housing provision and explore the benefits and challenges of undertaking longitudinal, mixed-methods research to address some of these difficulties. In using these methodologies, we faced a number of challenges around recruitment and retention, as well as strains on resources during both data collection and analysis. We outline how we dealt with these challenges and show how the use of a longitudinal, mixed-methods approach was crucial to producing a powerful dataset, which helped to answer a number of important questions.