Bearings are essential parts used in machine tools where high precision is required. It is important to understand bearing failure in order to replace a bearing before it affects the quality of precision. Bearing failure appears as a run-out in some applications, which is a critical factor for precision. However, previous studies have only focused on failure mechanisms, such as spalling. In this study, two types of wear models were used to predict the thrust bearings run-out: Linear and non-linear mechanisms. In order to validate the models, wear experiments of cross roller thrust bearings were performed. The average difference between the experiment and simulation run-out result was 16%. Then, the wear of different sized cross roller bearings was predicted by using a simulation. This was compared with the experiment result and showed up to a 6% difference. The suggested wear models are expected to be used to predict the failure/life of bearings.