There is controversy regarding the role of car mass in injury and fatality risk in individual car-to-car collisions and for overall car populations. In addition, the effect of frontal stiffness on both the case and partner cars is disputed, and the role of the distribution of collision closing speed has not been adequately examined. In this paper empirical car crash characteristics derived from onboard crash recorders are combined with risk functions based on mean vehicle acceleration. The model predictions closely match the available real life data for frontal collisions for the US (fatalities), Japan and Germany (AIS3+).

The model is used to predict the roles of case car mass, population car mass, collision closing speed distributions and frontal stiffness in relative vehicle safety. Results show that increasing the 50%ile car mass from 1200-1600kg reduces the AIS3+ injury to drivers by 4%, while decreasing the 50%ile car mass from 1200-800kg increases overall risk by 15%. By comparison, reducing the 99.99%ile collision speed from 200-150km/h reduces injury risk by 14%, but reducing the 50%ile collision speed from 60- 50km/h reduces injury risk by 31%. finally, a car population with the structural characteristics of the "soft" 50%ile reduces injury risk by 23%.