Background: Demographic variables (age and gender), automobile characteristics (weight, wheelbase, model) and driver behaviour (seat-belt use) can affect the outcome of a motor vehicle crash. Occupant fatality data suggest that frail elderly occupants have become more vulnerable than robust younger general traffic victims. In front impact, fixed object crashes, we hypothesized that greater age, female-gender, non-use of a restraining device, lower vehicle weight, shorter wheelbase and older models are positively associated with mortality risk. Methods: We used the U.S. Fatal Accident Reporting System (FARS) database to examine fatalities resulting from 12 o'clock fixed object collisions between 1986 and 1995. Our method mimicked, using real data, conditions under which crash test dummies are exposed to increased velocity of automobiles. Velocities at which fatalities occurred were then compared to driver/automobile characteristics using ANOVA/regression methods. The use/non-use of a three-point seat belt was also included in the analysis. Results: From 5,908 observations, age and gender were significantly related to travel speed at which a fatality occurred (p = .001). The estimated mean fatality travel speed for younger drivers was 63.7 mph, declining steadily to 46.3 mph for drivers aged ≥65. For male driver, fatalities, the mean travel speed was 59.6 mph compared to 53.1 mph for female drivers (p = .001). The age by gender interaction was significant (p = .015), although the gap between genders narrowed with increasing age. Other significant variables were weight of car, wheelbase and model year. The use of a seat-belt and the interactions between seat-belt use, age, and gender were not significant. Conclusions: Older adults and women are more susceptible to fatal injuries than younger adults and men. The frail elderly did not benefit from restraint use, as the use of seat-belts had no effect on the travel speed to fatally injure drivers. Inferences to non-fatal injuries cannot be made. Heavier vehicles and newer models had a protective effect, whereas wheelbase length was inversely related to travel speed. These data suggest that safety considerations cannot be dissociated from design issues. Future designs may have to consider the special needs of older adults and female drivers.