Introduction: Trauma care is highly complex and prone to medical errors. Accordingly, several studies have identified adverse events and conditions leading to potentially preventable or preventable deaths. Depending on the availability of specialized trauma care and the trauma system organization, between 10 and 30% of trauma-related deaths worldwide could be preventable if optimal care was promptly delivered. This narrative review aims to identify the main determinants and areas for improvements associated with potentially preventable trauma mortality. Methods: A literature review was performed using Medline, Embase and Cochrane Central Register of Controlled Trials from 1990 to a maximum of 6 months before submission for publication. Experimental or observational studies that have assessed determinants and areas for improvements that are associated with trauma death preventability were considered for inclusion. Two researchers independently selected eligible studies and extracted the relevant data. The main areas for improvements were classified using the Joint Commission on Accreditation of Healthcare Organizations patient event taxonomy. No statistical analyses were performed given the data heterogeneity. Results: From the 3647 individual titles obtained by the search strategy, a total of 37 studies were included. Each study included between 72 and 35311 trauma patients who had sustained mostly blunt trauma, frequently following a fall or a motor vehicle accident. Preventability assessment was performed for 17 to 2081 patients using either a single expert assessment (n = 2, 5,4%) or an expert panel review (n = 35, 94.6%). The definition of preventability and the taxonomy used varied greatly between the studies. The rate of potentially preventable or preventable death ranged from 2.4% to 76.5%. The most frequently reported areas for improvement were treatment delay, diagnosis accuracy to avoid missed or incorrect diagnosis and adverse events associated with the initial procedures performed. The risk of bias of the included studies was high for 32 studies because of the retrospective design and the panel review preventability assessment. Conclusion: Deaths occurring after a trauma remain often preventable. Included studies have used unstandardized definitions of a preventable death and various methodologies to perform the preventability assessment. The proportion of preventable or potentially preventable death reported in each study ranged from 2.4% to 76.5%. Delayed treatment, missed or incorrect initial diagnosis and adverse events following a procedure were commonly associated with preventable trauma deaths and could be targeted to develop quality improvement and monitoring projects.