Background: In many Emergency Department (ED) triage scoring systems, vital signs are not included as an assessment parameter. Objectives: To evaluate the validity of a new protocol for Emergency Medicine in a large cohort of patients referred to in-hospital care. Methods: From January 1 to June 30, 2006, 22,934 patients were admitted to the ED at Sahlgrenska University Hospital. Of those, 8695 were referred to in-hospital care and included in the study. A new five-level triage tool, combining vital signs, symptoms, and signs in the triage decision, was used. A small control of the inter-rater disagreement was also performed in 132 parallel, single-blinded observations. Results: Fifty percent of the patients were admitted by ambulance and the other 50% by walk-in. Hospital stay was significantly (p < 0.001) longer in those admitted by ambulance (9.3 +/- 14 days) as compared with walk-in patients (6.2 +/- 10 days). In-hospital mortality incidence was higher (8.1%) in patients admitted by ambulance, as compared with walk-in patients (2.4%). Hospital stay and in-hospital mortality increased with higher level of priority. In the highest priority groups, 32-53% of the patients were downgraded to a lower priority level after primary treatment. Conclusion: In the present study, the METTS protocol was shown to be a reliable triage method and a sensitive tool for secondary re-evaluation of the patient in the ED. (C) 2011 Elsevier Inc.