There is no method to measure dynamic frontal plane knee stiffness in humans while walking. Current testing methods, including in vivo dynamometer testing or ex vivo mechanical testing, exclude factors such as muscle forces and reflexes in weight-bearing environments. Quantifying dynamic stiffness could identify the presence or severity of knee joint injuries and inform practices to increase stability, protecting the knee from buckling unexpectedly. The purpose of this project was to design a tool to quantify frontal plane knee stiffness while walking. A wearable knee brace was developed to apply a dynamic knee abduction moment, via a medially-directed pneumatic perturbation force applied to the lateral epicondyle of the femur. Knee stiffness was estimated by measuring the applied load using a load cell and change in knee angle in the frontal plane using a synchronized optical motion capture system. Stiffness estimates from the brace were validated against a three-point bending test on a mechanical testing machine, using a sample of N=3 ovine cadaveric knees. Additionally, to test the brace’s ability to detect differences in stiffness between healthy and injured knees, the medial collateral ligaments of the same ovine knees were injured using a scalpel, and stiffness was measured again using both mechanical testing and brace methods. Knee stiffness was approximately 4 Nm/° for both brace and mechanical testing methods, but variance was much larger for the brace. Neither method was able to detect significant differences in knee stiffness between healthy and injured conditions, although a trend toward decreased stiffness for injured knees was observed in mechanical testing data when applied loads were below 2.5Nm. Further testing is required with more sensitive detection methods, more samples, and human subjects.