Abstract Background Patient and public involvement (PPI) is now an expectation of research funders, in the UK, but there is relatively little published literature on what this means in practice – nor is there much evaluative research about implementation and outputs. Policy literature endorses the need to include PPI representation at all stages of planning, performing and research dissemination, and recommends resource allocation to these roles; but details of how to make such inputs effective in practice are less common. While literature on power and participation informs the debate, there are relatively few published case studies of how this can play out through the lived experience of PPI in research; early findings highlight key issues around access to knowledge, resources, and interpersonal respect. This article describes the findings of a case study of PPI within a study about PPI in research. Methods The aim of the study was to look at how the PPI representatives’ inputs had developed over time, key challenges and changes, and lessons learned. We used realist evaluation and normalisation process theory to frame and analyse the data, which was drawn from project documentation, minutes of meetings and workshops, field notes and observations made by PPI representatives and researchers; documented feedback after meetings and activities; and the structured feedback from two formal reflective meetings. Findings Key findings included the need for named contacts who support, integrate and work with PPI contributors and researchers, to ensure partnership working is encouraged and supported to be as effective as possible. A structure for partnership working enabled this to be enacted systematically across all settings. Some individual tensions were nonetheless identified around different roles, with possible implications for clarifying expectations and deepening understandings of the different types of PPI contribution and of their importance. Even in a team with research expertise in PPI, the data showed that there were different phases and challenges to ‘normalising’ the PPI input to the project. Mutual commitment and flexibility, embedded through relationships across the team, led to inclusion and collaboration. Conclusion Work on developing relationships and teambuilding are as important for enabling partnership between PPI representatives and researchers as more practical components such as funding and information sharing. Early explicit exploration of the different roles and their contributions may assist effective participation and satisfaction.