Abstract: This study concentrates on the optimization of process parameters for surface roughness when turning hardened bearing steels using coated ceramic cutting tools and developed the multiple regression models for the surface roughness based on the factors cutting speed, feed rate, depth of cut and cutting fluid. The experimental results indicated that the feed rate was found to be a dominant factor on the surface roughness of Ra and Rz, but the other factors did not indicate the significant effect. In addition, optimal testing parameter for the surface roughness was determined. Average surface roughness obtained was about 0.79 and 4.102 μm for Ra and Rz, respectively, when machining the hardened steels. Furthermore, ANOVA revealed that the feed rate exerted a largest effect on the surface finish at contribution of 64.03%, followed by the depth of cut at 10.47%, respectively.