Background Fear of Falling (FoF) is commonly reported among older adults (up to 50%) and can impact negatively on physical and social activities, mood and quality of life. This paper explores the development, acceptability and feasibility of a cognitive behavioural therapy intervention (CBTi) for FoF. Methods The process evaluation of the CBTi development phase of an RCT (conducted in the UK) reported here, used ethnographic methods. Data included: interviews with patients and carers (n = 16), clinic staff (n = 6) and the psychologists developing the CBTi (n = 3); observational field notes and transcripts of intervention development meetings (n = 9) and stakeholder engagement meetings (n = 2); and informal discussions with staff developing the CBTi (n = 8). Data collection and thematic analysis were guided by Normalisation Process Theory (NPT). Results The process evaluation showed two domains of work necessary to develop a CBTi that made sense to stakeholders, and that could be delivered as part of an RCT. For the psychologists developing the content of the CBTi, a growing understanding of the complexity of FoF highlighted the need for an individualised rather than a manualised intervention. For the research team, the work involved adapting the structures and processes of the RCT to address preliminary concerns over the acceptability and feasibility of the proposed CBTi. Conclusions Theory-based approaches to process evaluation can sensitise researchers to contested understandings about proposed interventions that could undermine implementation. Drawing on the coherence construct of NPT, this study emphasises the nature and extent of work required to ensure an intervention makes sufficient sense to key stakeholders in order to maximise chances of successful implementation.