Purpose: Timely trauma care is crucial, especially for severely injured patients. We developed a set of criteria based on physiological changes, anatomic regions injured, and trauma mechanisms in our hospital to identify these patients. This study was performed to examine whether or not our revised activation protocol was strictly followed, and to identify areas of deficiency for ongoing quality improvement. The impact of the activation protocol on clinical outcome was also examined. Methods: We reviewed demographic data, mechanisms of injury, trauma scores, and outcome for all trauma patients admitted to a Level I trauma center in southern Taiwan between October 2010 and October 2011. All trauma protocol activation (TPA) decisions were subjected to careful scrutiny. Clinical outcomes for TPA patients and for non-TPA patients who fulfilled the activation criteria were compared. We also analyzed data for patients who did not meet our activation criteria but were found on subsequent work-up to have an Injury Severity Score (ISS) !16. Results: During Phase I of the study there were 2988 trauma admissions. The overall accuracy for TPA was 95.9%. Among 281 TPA patients, only 181 (64.4%) fulfilled the activation criteria. Among 204 patients who met the activation criteria, the protocol was not activated for 23 (11.3%). After we redesigned the computer program to prompt triage nurses and reeducated healthcare personnel, the false TPA and non-TPA rates improved to 14.6% and 2.8%, respectively, during Phase II of the study. There were 187 patients who did not meet our activation criteria and had ISS ! 16. TPA patients had a significantly lower Glasgow Coma Scale (GCS) score and were more likely to be admitted to the intensive care unit than non-TPA