Hybrid propulsion systems can give an effective contribution to compensate the low on-board energy storage capability of pure electric vehicles. The prediction of the performance of a hybrid thermalelectric vehicle is complex, due to the interaction of several components exchanging different kind of energy among them. A detailed model is required for the components selection, sizing and optimization, as well as for a model based control. This paper describes the theoretical and experimental activities developed at DOE for setting up a Hybrid City Minibus. A preliminary model validation has been carried out with in field data, evidencing the possibility to obtain a fully model based vehicle control.