The performance of the two point belt/knee bolster restraint and the three point belt restraint systems were compared based upon vehicle crash cases, computer simulations, and frontal sled tests. The National Accident Sampling System data base files for frontal impacts indicated that drivers restrained with two point belt/knee bolster systems experience significantly more liver injuries than occupants restrained with three point belt systems. To verify these findings, forty-four sled tests with human cadavers and the Hybrid III dummy were conducted at 32 km/h and 48 km/h. Subject instrumentation included upper and lower chest bands and thoracic accelerometers. Following the sled tests, radiographs and autopsy results were used to correlate cadaver injury with measured engineering parameters. Analysis of occupant kinematics using high speed films indicated greater longitudinal excursion of the hips and pelvis and smaller rotations of the torso for the two point belt/knee bolster restraint. The kinematic differences resulted in loading of the lateral chest and abdomen in the region of the liver for the two point belt/knee bolster system and loading of the upper chest for the three point belt system. Occupants with two point belt/knee bolster restraints incurred more liver and visceral injuries while occupants with three point belt restraints experienced more sternal and clavicular fractures.