In this study, a virtual dynamic friction test (DFT) framework was developed based on the finite element (FE) method to predict the coefficient of friction (COF) of pavement. Uniquely, this framework only necessitates the point cloud data of the target pavement, sourced via laser scanning, for input, and subsequently delivers COF-speed curves as output. A standout benefit is its foundation on the enhanced Persson’s friction theory for defining the interaction dynamics between the rubber slider and pavement. This contrasts with traditional FE-friction simulations, which rely on results from friction experiments. Consequently, preliminary friction tests are rendered redundant within our approach. During validation, the average predicted COF results for eight pavement samples, as determined by our framework, demonstrated no significant differences (with the number of parallel tests ¿ 6, p" role="presentation" style="font-size: 90%; display: inline-block; position: relative;"> p -value ¡ 0.05) compared to averages from actual DFTs. Offering parametric control over vital factors like water film thickness and rubber’s slip velocity, the virtual DFT model provides an innovative avenue to study the combined effects of water and velocity on pavement friction. Additionally, leveraging the virtual DFT methodology can slash the duration of field tests by over 50%.