The CRAMS scale and the Triage Revised Trauma Score (TRTS) were compared to assess their potential use as a prehospital method of activating hospital trauma teams. We studied patients from the resuscitation room of Leeds General Infirmary who had enough data recorded to allow calculation of the admission TRTS and CRAMS scale. Patients were defined as having major injury if they died in hospital, were admitted to the ICU or had an Injury Severity Score (ISS) of > 15. Each triage scale was compared by calculating multiple sensitivity/specificity pairs and plotting the results on a receiver operator (ROC) curve. The optimal cut-offs on each scale were compared directly. Ninety-seven (46 per cent) of a total of 213 patients fulfilled the study criteria for major injury. The best cut-off points were a CRAMS of < 9 and a TRTS of < 12. The TRTS was significantly more specific (0.9 versus 0.75) but at a cost of poor sensitivity (0.6 versus 0.69, not significant). The performance of both scales was similar when compared on the ROC curve. CRAMS and the TRTS were unable to identify major injuries in our sample with sensitivity and specificity adequate to support their use as a tool to activate trauma teams in the UK.