The majority of biomechanical tests that have been conducted to study knee injury tolerance and response in frontal crashes have used rigid or thickly padded flat-surfaced impactors that impact the anterior surface of the 90-degree flexed cadaver knee in a direction parallel to the long axis of the femur and without simulation of muscle tension. However, these impact conditions are not representative of typical knee-bolster-to-knee loading conditions in frontal crashes of today’s airbag-equipped vehicles. Laboratory tests are therefore being conducted to investigate the effects of knee angle and quadriceps muscle tension on knee response and knee injury tolerance as part of a larger study to characterize the response and tolerance of the knee, thigh, and hip to knee loading conditions that are representative of knee interactions with knee bolsters in frontal crashes of airbag-equipped vehicles. To determine representative knee angles and angles of the impactor surface relative to the femur, computer simulations of occupant kinematics in frontal impacts were performed using vehicle package geometry and occupant posture data from other UMTRI studies. In addition, tests were performed to determine the area over which forces are distributed across the knee during knee-to-knee- bolster interaction during frontal crashes by impacting Hybrid III ATD knees into the surface of production knee-bolsters. Based on these results, the impactor surface is angled at either 15 or 25 degrees from vertical and is padded to produce a contact area similar to that of the contact area between the Hybrid III knee and a production knee bolster. Isolated tibia-knee-femur specimens are prepared for testing by fixing the truncated ends of the femur and tibia in epoxy resin. The potted femur and tibia are then rigidly secured in the test fixture such that the long axis of the femur is horizontal and aligned with the direction of impactor motion. Pilot testing was conducted in which the effects of knee angle and quadriceps muscle tension on knee tolerance were explored while simultaneously exploring the effects of a knee-bolster-like loading condition on the response of the cadaver knee. Pre-impact tension in the quadriceps femoris tendon is set to either 5% or 50% of the estimated maximum force-producing capability of the thigh flexors by means of a cable parallel to the long axis of the femur attached to a force-limiting pneumatic actuator. In addition to measuring the applied force history, the response of each specimen was monitored using high-speed video and high-speed x-ray, where the latter provides information on patella kinematics and timing of knee fractures during impact loading. Preliminary results show that the type of knee fracture produced is related to the position of the patella at the time of fracture, with patella fractures occurring more often when the patella is above the center of the femoral condyles, and split condylar fractures occurring when the patella translates downward into the intercondyloid notch. Use of an impactor surface that is angled relative to the long axis of the femur increases the tendency for downward movement of the patella into the femora notch. However, the amount of downward patella movement is also affected by the initial patella position, which is a function of knee angle, and by the level of quadriceps muscle tension.