Condition monitoring of rolling element bearings is of vital importance in order to keep the industrial wheels running. In wind industry this is especially important due to the challenges in practical maintenance. The paper presents an attempt to improve the capability of prediction of remaining useful life of rolling bearings. The approach is based on the understanding of the wear of bearings i.e. wear modelling is briefly discussed. A simulation model has been built to produce vibration data of the monitoring of rolling bearings taking into account typical vibration excitations in addition to the wear. The simulation model is used to develop signal analysis methods and means of prognosis of the remaining useful life. One complete example of the above described process is shown and discussed in the paper.