OBJECTIVE: We sought to identify a combination of highly specific clinical variables that could be used to quickly identify a subset of high-need injured patients initially presenting to non-tertiary hospitals. METHODS: This was a retrospective cohort analysis of all injured adults 15 years or older meeting state trauma criteria, presenting to one of 42 non-tertiary hospital emergency departments (EDs) from January 1, 1998, through December 31, 2003, and surviving to ED disposition. The outcome included measures of timely resource need: early mortality (within 3 days of ED presentation), major nonorthopedic surgery within 3 days, or intensive care unit stay 2 days or longer. RESULTS: A total of 12,183 persons were included in the analysis, of which 3,643 (30%) patients had one or more of the outcome measures. The variables of greatest importance in identifying high-risk injured adults included (in order or priority): emergent airway intervention (prehospital or ED), initial ED GCS less than 11, ED blood transfusion, initial ED SBP less than 100 or more than 220 mmHg, and initial ED RR less than 10 or more than 32. These five variables had high specificity (89.1%, 95% confidence interval [CI] 88.2%-89.9%) in identifying 37.9% (95% CI 35.0%-40.7%) of high-risk trauma patients presenting to non-tertiary facilities. The positive likelihood ratio (+LR) for early mortality/early resource need increased for patients with one or more (+LR 3.5), two or more (+LR 9.1), and three or more (+LR 16.2) of the five risk criteria. CONCLUSIONS: There are five highly specific clinical risk criteria that may be useful in quickly identifying high-need injured persons presenting to non-tertiary hospitals. If validated, presence of these criteria may justify early higher level of care transfer by emergency medical services or mobilization of trauma resources without waiting for results of further diagnostic studies.