In this study, an enhanced analytical method is developed to obtain the energy loss during the loading-unloading interaction between roller bearing contact surfaces, in the presence of asperities. The macro-scale surface model considers the elastic-plastic interaction of the roller and outer ring. The micron-scale surface roughness parameters are defined based on three statistical parameters of the surface, which are the standard deviation of asperity height distribution, average asperity summit curvature and the density of asperities per unit area. The contact force induced by the interaction of rough surfaces is explicitly related to the minimum mean separation of surfaces. The energy loss in the bearing surfaces is then obtained using a numerical approximation and is related to the plasticity index, which is known to be a critical factor in the surface design of movement of the mechanical parts in contact. The developed model may assist manufactures to minimize the wear effect and risk of failure in roller bearing applications. Finally, for a lumped mass, assumed at the region of contact, the nonlinear vibration of the bearing system along with the identified contact force equation are utilized to estimate the contact natural frequency and damping ratio.