In (asphalt) fatigue the fatigue damage is related to tensile strains and stresses. I like to give another point of view. I "believe" that fatigue damage is firstly related to the Dissipated Distortion Energy (deformation of the shape of a unit volume) per load cycle and when the (Dissipated) Dilatation Energy (deformation of the volume of a unit volume) is positive (dVol>0) the fatigue damage is increased. When the volume change is negative (dVol 0 (so increasing fatigue damage). The same can be done with a linear visco-elastic multi-layer program e.g. VEROAD and the same yields for the dissipated energies. In the past using relatively thin pavements (h < 20cm) the main fatigue damage in the pavements was bottom-up cracking. Nowadays with relatively thick pavements (h > 30 cm) we are often faced with top-down cracking. Taking cores of damaged sections sometimes shows both types of cracks meeting each other halfway through the cores. In my view, the source of fatigue is directly related to the amount of dissipated distortion energy. For initiating the surface cracks the appearance of thermal stresses and tensile stresses at the edge of a tire are responsible but the material is already fatigued due to the distortion of the material in the top of the asphalt layer.