Knees of 13 unembalmed human cadavers and a Part 572 dummy were impacted with a pendulum having either 203 mm square aluminum honeycomb or Styrofoam DB striking surface. Impacts were made along the femoral axis of rigidly mounted legs with a 52.3 kg pendulum at velocities of 1.8 and 3.6 m/s. Forces transmitted through the femur and tibia (at the ankle) were measured to determine the force response distribution. Pendulum acceleration parallel to the femoral axis was measured for all tests. Knee penetrations into the striking surface material were obtained by double integration of the pendulum acceleration-time traces, were verified by analysis of high speed movies, and were compared with the projected contact areas as penetration increased. Impact responses are presented as crossplots of force versus depth of knee penetration into the interface material.For aluminum honeycomb impacts at 3.6 m/s, the mean peak force resulting from cadaver impacts is lower than the mean peak force resulting from dummy knee impacts. However, for Styrofoam DB impacts at 3.6 m/s, the mean force-penetration response of the cadaver knees is similar to the response of the Part 572 dummy knee. Significant differences in peak force levels for the cadaver and dummy were not observed with either aluminum honeycomb or with Styrofoam DB at 1.8 m/s impact velocities.