The biodynamic response of the femur during passively restrained -Gx impact acceleration is reported in this paper. Eleven unembalmed cadavers, ranging in age from 21 to 65 and weighing from 50 to 96 kg, were tested in a VW Rabbit seat with a passive belt and knee restraint. Sectioned parts of the VW knee bolster were placed about 130 mm away from the patella at the initiation of the tests. The height of the knee bolsters was adjusted individually in the eleven tests. Ten were set for loading directly through the patella. In one run, the impact was below the knee joint. The sectioned bolsters were mounted on a rigid frame and instrumented with triaxial load cells. A six-axis load cell was installed in the right femur. Photo targets were attached directly to the femur and tibia.Sled runs were made at 22 and 35 g. Only one cadaver sustained bilateral femoral fractures at 35 g. Bolster loads were twice as high as those reported previously and femoral loads averaged 65% of the measured bolster load. There was only one case of avulsion of the posterior cruciate ligament for impact knee angles ranging from 83 to 119 deg. The shape of the bolster prevented initial tibial contact and lessened the risk to this ligament. Bending moments along the femoral shaft were found to be high. They should be considered in any new injury criteria for the femur.