Various architecture designs exist for hybrid and plug-in hybrid vehicle powertrains. Through-The-Road (TTR) layout is an interesting solution and is adopted in some vehicles. In TTR, both front and rear axles provide power to the wheels: one is associated to the engine traction system while the other is linked to the electric traction system. Different design variants of TTR exist. They differ in their fuel saving benefits and their required component sizing which affects the powertrain cost. In this paper, a general TTR powertrain architecture is considered and modelled. Different design variants are then optimized using a bi-level sizing and control optimization methodology. It is based on the genetic algorithm approach for the choice of sizing, combined with a global optimal energy management strategy (Dynamic programming) for the calculation of the first objective function, the fuel consumption. The second objective function is a powertrain cost index. The aim of the paper is to determine the trends in design modification and sizing variation that can allow minimizing the powertrain cost and fuel consumption.