BACKGROUND: Attempts to design services to support the delivery of healthcare closer to home have taken various forms as countries respond to an increase in hospital admission rates for older people, who are at risk of hospital-acquired morbidity, prolonged lengths of stay and readmission. Evidence to support the development of these services is limited. We are conducting a process evaluation, alongside a UK multi-site randomised trial, to understand the contexts and practices of implementing geriatrician-led admission avoidance hospital at home services and to explore ways that the intervention might be effective, under what conditions, for whom, and how it differs from inpatient care. METHODS: We are interviewing patients and their caregivers, from sites that are purposively sampled from participating National Health Service (NHS) trusts across the UK. We are also visiting sites to observe local processes and discuss the establishment and running of services with a range of multidisciplinary staff, managers, commissioners, primary care and social services representatives. We aim to interview approximately 36 patients and their caregivers with experience of hospital at home or inpatient services; 12 at each of three sites. We will use a content analysis approach to explore data across participants, services and sites. DISCUSSION: This process evaluation will enable evaluation of implementation processes prior to knowing trial outcomes. We encompass domains of reach, delivery, change, context and response to the intervention by patients, their carers, health professionals and the health system. TRIAL REGISTRATION: ISRCTN60477865 . Registered on 10 March 2014. Trial sponsor: University of Oxford. Version 3.1, registered on 14 June 2016.