Numerous researches about hybrid electrical vehicles (HEV) deal with topologies, technologies, sizing and control. These aspects allows to reduce transportation cost and environmental impacts. The paper focuses on the modelling of the electrical motor of the HEV in a global sizing context, taking into account: the hybrid system, the driving cycle and an optimal energy management. In the paper, the parallel hybrid electrical vehicle (HEV) will be our study case. In a classical HEV design process, scaling model is usually used to fix the standard power of the electrical machine. The efficiency and the maximum torque power are scaled using a linear dependency on the rated maximum power. The paper compares two sizing models: a scaling model based on a efficiency map and a scaling model based on a magnetic circuit model (MCM). A comparison is performed using both models in an optimization process. This optimization process is a global sizing process using dynamic programming as an optimal energy management. Optimal sizings of the hybrid vehicle are compared, depending on whether they use a classical sizing method or a MCM scaling method.