Effect of roughness orientation on lubricant film thickness has been an important issue of surface design, attracting much attention since the 1970 s. A systematical study, however, is still needed for various contact types in an extended range of operating conditions, especially in mixed lubrication cases with film thickness to roughness ratio (λ ratio) smaller than 0.5. The present study employs a deterministic mixed elastohydrodynamic lubrication (EHL) model to investigate the performance of lubricating films in different types of contact geometry, including the line contact, circular contact, and elliptical contacts of various ellipticity ratios. The speed range for analyzed cases covers 11 orders of magnitude so that the entire transition from full-film and mixed EHL down to dry contact (corresponding λ ratio from about 3.5 down to 0.001 or so) is simulated. Three types of machined surfaces are used, representing transverse, longitudinal, and isotropic roughness, respectively. The line contact results are compared with those from the stochastic models by Patir and Cheng (“Effect of Surface Roughness Orientation on the Central Film Thickness in EHD Contacts,” Proc. 5th Leeds-Lyon Symp. on Tribol., 1978, pp. 15–21) and the influence of roughness orientation predicted by the deterministic model is found to be less significant than that by the stochastic models, although the basic trends are about the same when λ > 0.5. The orientation effect for circular or elliptical contact problems appears to be more complicated than that for line contacts due to the existence of significant lateral flows. In circular contacts, or elliptical contacts with the ellipticity ratio smaller than one, the longitudinal roughness may become more favorable than the isotropic and transverse. Overall, the orientation effect is significant in the mixed EHL regime where theλratio is roughly in the range from 0.05 to 1.0. It is relatively insignificant for both the full-film EHL (λ > 1.2 or so) and the boundary lubrication/dry contact (λ < 0.025 ∼ 0.05).