The lock mechanism is one of the key mechanisms in the aircraft. Wear of the hinge lead to an increase of the clearance and a decrease in matching accuracy, and the kinematic accuracy of the lock mechanism decreases or even the mechanism may fail. However, the existing hinge prediction methods mainly focused on the physic-based model without utilizing condition monitoring data. The physical model cannot update the contact condition timely because of the contact condition in the hinge varies with time. The disadvantage of the physical model is that the predicted errors are growing with time. To solve this problem, a new approach is presented in this work. Firstly, a dynamic model of the lock mechanism is built, and the load case of the hinge is obtained. Archard's wear law is used to analyse the wear of it. In this model, the wear coefficient is related to the contact condition of the hinge. If a more accurate wear coefficient can be obtained, the evolution of wear will be predicted more accurately. To achieve this, a Bayesian update process is implemented to incorporate the wear depth observation at an inspection point to determine the posterior distribution of the wear coefficient. The prediction approach is validated by the experiment data of the hinge of the lock mechanism.