Two proposed methods to determine the adhesion friction coefficient were validated by experimental results of two types of rubber compounds at different sliding velocities under dry conditions. The experimental results were measured from a linear friction tester, while the viscoelastic friction coefficient was estimated using the Persson's contact theory. Adhesive friction (model 1) was derived from the deconvolution of dry friction coefficient in two Gaussian-like curves. Interesting results were obtained using the deconvoluted method in the range of intermediate sliding velocities where preponderant contribution to the adhesion friction is replaced by the viscoelastic friction. Fitting parameter results were in good general agreement with values derived from the literature, confirming the influence of the mechanical properties of the compound and substrate texture on the proposed adhesion frictional method. The second adhesive friction model (model 2) was based on the confinement rheology of rubber chains on the contact with the asperities of the road surface. We demonstrated that acceptable adhesion friction results were achieved from a dynamic viscosity test at low frequencies, confirming the applicability of the proposed rheological model. Moreover, the relationship between the rubber composition and the modified contact layer along with the likely interphase reaction are also discussed.