The TRISS methodology for evaluating trauma care has been applied in two UK hospitals over a 12 month period. Unexpected deaths occurred in both hospitals and were related to delays in transport between hospitals. The delays were instigated by the receiving hospitals and never the despatching hospital. Unexpected survival occurred in one of the hospitals. The common factors were: 1. Patients were taken there directly and did not require further transfer. 2. They were resuscitated by a consultant. 3. They were electively ventilated. 4. CT Scan excluded a neurosurgical emergency. 5. When surgery was required it was carried out early. 6. They were treated for several weeks in an ICU with 24 hour direct consultant involvement. They all developed septic shock and acute respiratory distress but ultimately made a full recovery. This evidence supports the view that direct transfer of patients to Trauma Centres in the UK is likely to reduce unnecessary deaths. The development of such a system and the location of such centres must be based on objective measures of case mix and performance. We recommend the establishment and development of a Major Trauma Outcome Study (MTOS) (UK).