Objective: In a Trauma System, Trauma Registry allows the assessment of epidemiology and quality of patient care. Materials and Methods: Data about trauma patients admitted to Ospedale Niguarda Emergency Department from October 1, 2002 to June 30, 2003 with ICD9CM code 800-939.9 and 950-959.9 were prospectively recorded. Injury severity score (ISS) and revised trauma score (RTS) were calculated and probability of survival (Ps) was derived. Results: During the study period 1811 trauma patients were admitted, and 271 (14.96%) were consistent with triage criteria of severity. Among these, survivors were 220 (81.18%) and blunt trauma 95.94%. Injuries of the central nervous system with (11.76%) or without (50.98%) hemodynamic instability or hemodynamic instability alone (31.37%), were the principal causes of death. An ISS greater than 15 was observed in 61.25% with an overtriage of 38.75%. Seventy seven patients accepted without triage criteria of severity died or were admitted to intensive care unit with an undertriage of 5%. Ps among victims was 22.35±27.19 and possible preventable deaths were 6 (11.76%). No frankly preventable death was recorded. Discussion: Standard pre-hospital triage criteria are associated with significant over and undertriage. Data collection using large population-based data base increases epidemiologic value of trauma registry. Analysis of Ps identifies cases who need clinical discussion to assess adequacy of treatment. Conclusions: Prospective data collection in a trauma registry may provide all informations useful to improve quality of trauma patient care.