Background: The city of Shenzhen, China, is planning to establish a trauma system. At present, there are few data on the geographical distribution of incidents, which is key to deciding on the location of trauma centres. The aim of this study was to perform a geographical analysis in order to inform the development of a trauma system in Shenzhen. Methods: Retrospective analysis of trauma incidents attended by Shenzhen Emergency Medical Services (EMS) in 2014. Data were obtained from Shenzhen EMS. Incident distribution was explored using dot and kernel density estimate maps. Clustering was determined using the nearest neighbour index. The type of healthcare facilities which patients were taken to was compared against patients’ needs, as assessed using the Field Triage Decision Scheme. Results: There were 49,082 recorded incidents. A total of 3513 were classed as major trauma. Mapping demonstrates that incidents predominantly occurred in the western part of Shenzhen, with identifiable clusters. Nearest neighbour index was 0.048. Of patients deemed to have suffered major trauma, 8.5% were taken to a teaching hospital, 13.6% to a regional hospital, 42.6% to a community hospital, and 35.3% to a private hospital. The proportions of Step 1 or 2 negative patients were almost identical. Conclusion: The majority of trauma patients, including trauma patients who are at greater likelihood of severe injury, are taken to regional and community hospitals. There are areas with identifiable concentrations of volume, which should be considered for the siting of high-level trauma centres, although further modelling is required to make firm recommendations.