This paper presents results of a research aiming at modeling the evolution of road surface skid resistance using statistical and phenomenological approaches. Data are provided by a database composed of measurements collected on French roads during ten years: skid resistance values (Sideway Force Coefficient measured by SCRIM machine), macrotexture values (Mean Profile Depth), road geometry, traffic and age. Principal component analyses determine the connection between relevant factors. Analyses of variance enable the selection of independent variables used in the evolution laws. Various types of evolution laws are explored: linear regressions with one to four parameters, non-linear regressions and logarithmic functions. In parallel, model based on the combination of the main factors that can explain the skid resistance evolution is developed. This model takes into account both mechanical effect like polishing under traffic and ageing effect (changes in asphalt mix properties due to the weather effect). In a last part of the study, results obtained with the two approaches on a specific database are compared with results based on models found in literature review.