Abstract: Electrostatic sensor is used to investigate scuffing experiments of oil lubricated sliding bearing steel with point contacts. The charging mechanism and electrostatic induction principle are analyzed. Then the scuffing experiments are carried out, and the denoising method of the electrostatic signals is studied. Finally the influence of the load and speed on the friction coefficient and the electrostatic signals are discussed. The results show that the difference spectrum theory of singular value is suitable for the electrostatic signals denosing. The electrostatic technology (EST) and friction coefficient could be used to detect the occurrence of scuffing. In addition, EST could detect the precursor signals before the occurrence of scuffing, and this is a good support for the condition based maintenance. The amplitude and root mean square (RMS) of electrostatic signal embody four stages (early wear, stablel wear, early scuffing and scuffing). In the stable wear stage, the friction coefficient increases as the load and speed decrease. RMS is not sensitive to the varying load and it increases with increasing speed.