Background: In Italy there isn't a State Trauma System. Many attempts have been done to increase the quality of trauma care in prehospital and hospital phases, but only by local resources. In Mila (Italy)o Emergency Medical System is organized by Regional rules and five Hospitals warrant high level of care for trauma patients. There isn't an official registry for trauma. Creating a Trauma Registry is the prerogative to analyse the quality of assistance and to propose new solutions., Objectives: To analyse major trauma patients admitted to Ospedale Maggiore Policlinico IRCCS; to evaluate diagnostic and therapeutic protocols in order to identify preventable deaths., Patients: We have observed trauma patients admitted to Ospedale Maggiore from January to December 2004. We collected demographic data, informations about the traumatic event and prehospital rescue, emergency room examination, diagnostic exams, surgical operations and results of treatment. We selected patients admitted among 6917 trauma patients observed in this period. We have calculated RTS, ISS and TRISS. Patients were followed during their staying at the hospital to record length of staying, lenght of ICU and mortality rate. We collected the autopsy of the all death patients., Results: We selected 299 patients, 207 males and 92 females. Mean age was 42.4 +/- 19.5 for males (range 15 - 99) and 57.7 +/- 22.5 for female (range 7 - 101). Motorvehicle and road incident were the main cause of trauma (55.5%). A penetrating injury was observed only in 5% of cases. Mean RTS was 7.5 +/- 1. ISS and TRISS were (mean +/- SD) 13 +/- 9 and 94.9% +/- 11.5, respectively. Patients with ISS = 16 were 109 (36.4%). Forty five patients (15%) required a surgical treatment during the first 48 hours. Total length of staying was 8.9 +/- 11.2 (mean +/- SD) days (median of 5.5 days) and the length of ICU was (mean +/- SD) 11.7 +/- 10.3 days (median 9 days). 12 patients died (mortality rate 4.08%), 11 at Policlinico (2 in the emergency room, 3 in the operative room, 5 in ICU. One patient died in surgical ward), 1 at Ospedale Niguarda. Autopsy was available for 8 patients. In 2 cases the cause of death was established by clinical examination and in 1 case police are still investigating for poisons or other letal drugs. The main cause of death was the cerebral injury. Only for 1 patient it was impossibile to determine the cause of death so he was considered a potentially preventable death. His clinical RTS in the emergency room was 12 (7,4808 in the statistical analysis) and no severe lesions were observed during primary and secondary survey., Conclusions: Our data are typical of an urban area of a western country. Penetrating injury are very rare, 5% of incidence. Diagnostic and therapeutic protocols are similar to countries where a Trauma Center is active. The 4% of overall mortality rate is similar to Trauma Centers in USA. This result is better than other hospitals in Milan. The high number of ATLS providers in the trauma team could be one of causes of good results. Quality audit can't consider only RTS, ISS and TRISS. Scores are very practical and useful but they aren't enough. We must analyse every single case of death and Trauma Registry is the first tool to evaluate trauma care in a modern EMS.