BACKGROUND: Palliative care is a policy priority internationally. In England, policymakers are seeking to develop high quality care for all by focusing on reducing the number of patients who die in acute hospitals. It is argued that reducing 'inappropriate' hospital admissions will lead to an improvement in the quality of care and provide cost savings.Yet what is meant by an 'inappropriate' admission is unclear and is unlikely to be shared by all stakeholders. The decision process that leads to hospital admission is often challenging, particularly when patients are frail and elderly. The ACE study reopens the idea of 'inappropriate' hospital admissions close to the end of life. We will explore how decisions that result in inpatient admissions close to death are made and valued from the perspective of the decision-maker, and will consider the implications of these findings for current policy and practice. DESIGN/METHODS: The study focuses on the admission of patients with advanced dementia, chest disease or cancer who die within 72 hours of admission to acute hospitals. The study uses mixed methods with three data collection phases. Phase one involves patient case studies of admissions with interviews with clinicians involved in the admission and next-of-kin. Phase two uses vignette-based focus groups with clinical professionals and patients living with the conditions of interest. Phase three uses questionnaires distributed to clinical stakeholders. Qualitative data will be explored using framework analysis whilst the questionnaire data will be examined using descriptive statistical analysis. Findings will be used to evaluate current policy and literature. DISCUSSION: Significant ethical and validity issues arise due to the retrospective nature of phase one of the study. We are not able to gain consent from patients who have died, and the views of the deceased patients cannot be included directly, which risks privileging professional views. This phase also relies on the memories of the participants which may be unreliable. Later phases of the study attempt to compensate for the "absent voices" of the deceased patients by including next-of-kin and patient focus groups.