Regarding the two different contributions to the driving force, the direct force and the wind force, the role of the electronic structure has been quite different for the two. For the wind force increasingly sophisticated descriptions have been used, namely pseudopotential models, finite cluster models and, at the end, an ab initio Korringa-Kohn-Rostoker (KKR) Green's function description. We will illustrate this by showing for which systems by now the wind force has been calculated, which include almost all Fee and Bee metals, while both self-electromigration and impurity migration have been treated. Some new results will be presented as well, which simulate electromigration along a grain boundary and over a surface.The direct force, on the other hand, has mainly been discussed in terms of the simple free electron, or jellium model. However, it will be shown that we have arrived at a point, at which more sophisticated descriptions of the electronic structure involved are becoming important. A recent analysis of new experimental results leads to the conclusion that a migrating hydrogen atom effectively can have a direct valence smaller than unity, depending on the metal studied. By this it becomes challenging to perform calculations of the electronic structure of an interstitial, not only at its equilibrium position, but also at positions lying along the jump path.