A procedure is presented to estimate the risk to life that multiple injuries pose to crash victims. It continues Eppinger’s original work first presented at the 1981 ESV conference. At its core, the procedure uses only the two most serious injuries – denoted as the primary injury and the secondary injury – to characterize a victim’s entire injury record. Nine years of data from the Crashworthiness Data System (CDS) containing injury records for over 50,000 crash victims – including over 3500 fatalities – are analyzed. For each victim, the top two injuries are defined by using a data-driven approach based on actual CDS outcomes as opposed to relying solely on the Abbreviated Injury Scale, a heuristic ranking system developed by a panel of experts. Results show that for a given primary injury, the risk to life varies profoundly depending upon the secondary injury. Victim age has a substantial effect, too. When deviance statistics are considered, the new procedure predicts fatalities better than other injury scales (including the Injury Severity Score). Ultimately, this two-injury procedure promotes better estimates of safety benefits by directly quantifying and specifying fatality-related injuries in the CDS data.