UNLABELLED Most people who die in hospital do so with a DNACPR order in place, these orders are the focus of considerable debate. AIM To identify factors, facilitators and barriers involved in DNACPR decision-making and implementation. METHODS All study designs and interventions were eligible for inclusion. Studies were appraised guided by CASP tools. A qualitative analysis was undertaken. DATA SOURCES Included electronic databases: Medline, Embase, ASSIA, Cochrane library, CINAHL, PsycINFO, Web of Science, the King's Fund Library and scanning reference lists of included studies. RESULTS Four key themes were identified: Considering the decision - by senior physicians, nursing staff, patients and relatives. Key triggers included older age, co-morbidities, adverse prognostic factors, quality of life and the likelihood of success of CPR. Discussing the decision - levels, and combinations, of physician and nursing skills, patient understanding and family involvement produced various outcomes. Implementing the decision - the lack of clear documentation resulted in a breakdown in communications within health teams. Staff knowledge and support of guidelines and local policies varied. Consequences of a DNACPR decision - inadequate understanding by staff resulted in suboptimal care, and incorrect withdrawal of treatment. CONCLUSION Significant variability was identified in DNACPR decision-making and implementation. The evidence base is weak but the absence of evidence does not indicate an absence of good practice. Issues are complex, and dependent on a number of factors. Misunderstandings and poor discussions can be overcome such as with an overall care plan to facilitate discussions and reduce negative impact of DNACPR orders on aspects of patient care.