Although scores and other prehospital triage schema effectively identify injured patients who will benefit from trauma center care, those tools are relatively nonspecific. One consequence is overtriage--transport of less severely injured patients to trauma centers--with resulting expenditure of scarce resources on patients who do not benefit from an emergent and intensive response. We developed a tool that, during the prehospital phase, can sort inner-city trauma victims into those who will require ICU/OR services and those who will not. Beginning with the trauma triage criteria promulgated by the American College of Surgeons, we created two tiers of personnel response based on vital signs, mechanism of injury, and limited anatomic information ordinarily reported by prehospital providers. Based on our initial experience with the two-tier response, the sorting criteria were revised and refined. The sensitivity of the current version of the two-tier criteria for predicting which trauma patients will require ICU/OR services during the first 24 hours of hospitalization approaches 95% (excluding misapplications of the tool) while avoiding urgent trauma team mobilization in 57% of patients triaged to our trauma center. Two-tiered trauma responses appear to be safe and may represent an important strategy for more effective distribution of increasingly scarce and costly resources.