Pavement surface friction significantly impacts road safety and is predominantly influenced by pavement texture characteristics. Traditionally, mean texture depth (MTD) has been widely used as a primary indicator for friction prediction. However, reliance solely on MTD neglects crucial surface texture variables such as exposed aggregate number (EAN) and aggregate shape, potentially limiting prediction accuracy. This study evaluates the field applicability of a pavement surface friction prediction model previously developed using 3-D–printed specimens, incorporating MTD, EAN, and aggregate shape as key predictors. Field data from 184 pavement sites, including exposed aggregate concrete pavement, dense graded asphalt concrete, and open graded asphalt concrete, were analyzed. Measurements of surface texture (MTD, EAN) and friction British pendulum number were collected, and the aggregate shape index was optimized for real pavement conditions. The application of the 3-D–printed specimen-based prediction model to actual pavement surfaces yielded an R² of 51 % with a P value of 0.00, demonstrating moderate predictive capability. The analysis identified an optimal aggregate shape index (S = 2.75), indicating that aggregate shapes between spherical and pyramidal significantly influence pavement friction. Moreover, the study confirmed a nonlinear relationship between EAN and friction, with an optimal EAN around 30 for maximum friction. These findings highlight the importance of incorporating additional texture variables beyond MTD for improved accuracy in pavement friction prediction. This research provides a foundation for more comprehensive and effective models, contributing to enhanced pavement safety management.