Modeling the real contact area plays a key role in every tribological process, such as friction, adhesion, and wear. Contact between two solids does not necessarily occur everywhere within the apparent contact area. Considering the multiscale nature of roughness, Persson proposed a theory of contact mechanics for a soft and smooth solid in contact with a rigid rough surface. In this theory, he assumed that the vertical displacement on the soft surface could be approximated by the height profile of the substrate surface. Although this assumption gives an accurate pressure distribution at the interface for complete contact, when no gap exists between two surfaces, it results in an overestimation of elastic energy stored in the material for partial contact, which typically occurs in many practical applications. This issue was later addressed by Persson by including a correction factor obtained from the comparison of the theoretical results with molecular dynamics simulation. This paper proposes a different approach to correct the overestimation of vertical displacement in Persson’s contact theory for rough surfaces with self-affine fractal properties. The results are compared with the correction factor proposed by Persson. The main advantage of the proposed method is that it uses physical parameters such as the surface roughness characteristics, material properties, sliding velocity, and normal load to correct the model. This method is also implemented in the theory of rubber friction. The results of the corrected friction model are compared with experiments. The results confirm that the modified model predicts the friction coefficient as a function of sliding velocity more accurately than the original model.