As one of the most common parts of various rolling mechanical equipments, rolling element bearing is vulnerable. Therefore, great attentions have been attributed to the theories, failure diagnosis methods and their applications for rolling bearings. Vibration analysis is also a very important means for bearing fault diagnosis. This paper aims at the research on the methods of signal processing and pattern recognition. An experimental platform was set up for the failure diagnosis of rolling bearings, on which we have done a lot of experiments. Then the vibration signals of normal rolling bearings, rolling bearings with failure on the outer and inner race were collected. Time-delayed correlation demodulation was applied and the features of vibration signal were effectively extracted. Fuzzy C-means clustering system was established to carry out the recognition of the fault of bearings. Experimental results have proved the developed fault diagnostic architecture is reliable and effective.