Most of the analyses are based on fatality analysis reporting system (FARS) data from 1988 through the first six months of 1997. The primary analysis compares the fatality risk for back seat outboard belted occupants (lap or lap/shoulder belted) to the corresponding risk for unbelted occupants, as well as the fatality risk for lap/shoulder belted occupants to the risk for lap belted occupants. Fatality risk is the ratio of fatalities in the back seat to fatalities in the front seat (a control group). This procedure of comparing a subject group to a control group is called "double pair comparison." The principal conclusions are: back seat lap belts are 32 percent effective in reducing fatalities and lap/shoulder belts are 44 percent effective in reducing fatalities when compared to unrestrained back seat occupants in passenger cars. In passenger vans and sport utility vehicles, lap belts are 63 percent effective and lap/shoulder belts are 73 percent effective. The change from lap to lap/shoulder belts has significantly enhanced occupant protection, especially in frontal crashes. In all crashes, lap/shoulder belts are 15 percent more effective than lap belts alone. In frontal crashes, lap/shoulder belts are 25 percent more effective than lap belts alone. Back seat lap belts reduce the risk of head injuries while increasing the risk of abdominal injuries in potentially fatal frontal crashes. Lap/shoulder belts reduce the risk of both head and abdominal injuries in potentially fatal frontal crashes relative to lap belts only: head injuries by 47 percent and abdominal injuries by 52 percent.