Objective: Australians are more likely to die in a hospital than anywhere else, and most of these deaths are 'expected'. The aims of this survey were to identify if specific end-of-life issues were documented in clinical records of a regional hospital serving remote Australia. Design: A retrospective consecutive case-note audit covering 18months from 1 January 2006. Setting: Alice Springs Hospital. Participants: Adult deaths in Alice Springs Hospital. During this period there were 128 deaths. Outcome measures: Demographic, process (diagnosis of dying documented, 'not for resuscitation' documentation, intensive care unit (ICU) admission, social worker referral and specialised palliative care service referral) and outcome data ('comfort at end of life') were surveyed. Results: Of the 128 admissions, 55 cases were excluded: 10 were children under 18, 33 died in <48 hours, nine were coroner's cases and three files could no't be found. Of the 73 deaths surveyed (33 men), 47 (64%) were Aboriginal. A diagnosis of dying was made in 84%, 88% had an 'not for resuscitation' order, 48% were admitted to ICU during their last admission, 66% were referred to social workers, 68% were referred to palliative care and 85% of people were documented to be 'comfortable' during the dying process with no differential outcomes for Aboriginal and non-Aboriginal decedents except age (P<0.0001). Conclusion: This survey highlights the continuing need to diagnose dying, understand optimal use of intensive care and improve comfort care at the end of life. �� 2011 The Authors. Australian Journal of Rural Health �� National Rural Health Alliance Inc.