A critical point for accurate prediction of asphaltic pavement performance life is to model as many sources of energy dissipation as possible, such as viscoelastic constitutive behavior and crack propagations. It is known from experimental observation that many cracks can occur simultaneously in a pavement structure and that these cracks can coalesce into a single crack that leads to structural failure. A multiscale approach accounts for multiple cracks and provides computational efficiency. This study presents a multiscale computational model for predicting evolution of damage in asphaltic pavements. Two scales are included in the analysis, termed: global scale (pavement structure) and local scale (RVE of asphalt concrete mixtures). The model produces damage-dependent behavior of pavements by linking the properties of mixture constituents (local scale) and pavement structural performance (global scale). Damage is included in the forms of cracks modeled with viscoelastic cohesive zones. The model uses material properties at the constituent level (local scale) as inputs, and a FEM time-marching algorithm. It performs simultaneous computations on both local and global scale. Preliminary stage of the modeling at this time emphasizes computational results depending on different material properties of mixture constituents to monitor the impact of different materials on asphaltic fracture behaviour and pavement failure.