To address the current lack of quantitative performance degradation models for elevator brakes, this paper conducts performance degradation experiment and modeling studies on elevator brakes. By establishing a simulated experimental platform for elevator braking performance, 1 500 braking tests were conducted on the same set of brake friction pads under full-speed and full-load conditions to obtain experimental data on braking torque degradation. The Gamma process was used to simulate the degradation trend of braking torque, while the random relationship between observed braking torque and trend values was analyzed using a normal distribution. A braking performance degradation prediction model was then constructed based on monte carlo simulation. The results show that the Gamma distribution can effectively characterize the statistical law of wear increment. The degradation prediction model constructed based on this law fits well with the degradation trend of the experimental data in this paper, providing an effective methodological reference for the performance degradation research of elevator brakes.