Pavement surface friction is a significant factor for driving safety and plays a critical role in reducing wet-pavement crashes. However, the current asphalt mixture design procedure does not directly consider friction as a requirement. The objective of this study was to develop a surface friction prediction model that can be used during a wearing course mixture design. To achieve the objective, an experimental study was conducted on the frictional characteristics of typical wearing course mixtures in Louisiana. Twelve wearing course mixtures including dense-graded and open-graded mixes with different combinations of aggregate sources were evaluated in laboratory using an accelerated polishing and testing procedure considering both micro- and macro texture properties. In addition, the surface frictional properties of asphalt mixtures were measured on twenty-two selected asphalt pavement sections using different in situ devices including Dynamic Friction Tester (DFT), Circular Texture Meter (CTM), and Lock-Wheel Skid Trailer (LWST). The results have led to develop a procedure for predicting pavement end-of-life skid resistance based on the aggregate blend polish stone value, gradation parameters, and traffic, which is suited in checking whether the selected aggregates in a wearing course mix design would meet field friction requirements under a certain design traffic polishing.