After patients have a total knee replacement, they lose the ability to flex their knees past about 120°. This makes activities like squatting, kneeling to pray, and getting up or down from the floor very difficult, if not impossible. Our research attempts to understand why current total knee replacement designs limit the amount of knee flexion by simulating deep flexion activities in a cadaveric model. The simulations are performed using the Kansas Knee Simulator, a servo-hydraulic loading rig that recreates dynamic activities on a cadaveric knee like walking or climbing stairs. By testing on a cadaveric knee, we can evaluate the natural motion of the knee during a battery of activities then perform total knee replacements to see how those motions change. The direct comparison between the natural and surgically repaired knee give insight into the behavior of the knee implant and how factors like implant alignment, ligament quality, or surgical error will affect the patient’s outcome. In addition to influencing future implant designs, this research is also being used to build and validate computational models to predict knee behavior before and after total knee replacement. The ultimate goal of these computational models is to be used during surgery to help surgeons optimize their surgical technique to maximize the performance and longevity of the implants.