There is continuing controversy regarding the role of car mass in occupant – risk. Injury risk is influenced not only by the vehicle deformation characteristics and the occupant restraint system, but also by the size and mass of both case and partner cars. Recent research has shown that injury is better correlated with mean car acceleration than with velocity change. Injury risk relations for belted/unbelted drivers in all crashes with AIS 3+ and fatal injuries were derived from US data using a modified Joksch [7] type risk equation. These risk equations are shown to compare with recent Swedish AIS 2+ and AIS 3+ frontal collision data and with US frontal collision fatality data. The derived AIS 3+ and fatality risk functions, in conjunction with car population stress/density characteristics, were combined with US, German and Japanese car mass and collision velocity distributions to predict the variation in fatality and AIS 3+ injury risk with car mass and mass ratio using Monte Carlo simulation techniques. The resulting predictions were compared to published real world accident risk versus car mass data for each country, and a high degree of correspondence obtained. The results show that empirically derived car population structural characteristics and injury risk functions satisfactorily explain the real world fatality and injury versus mass effect.