Nowadays, the design of Power Modules (PM) is shared between multidisciplinary teams of designers using various modelling tools. Consequently, optimizing the layout of PM is a long process with high risk of errors.In this paper, a sequentially coupled approach is proposed to optimize this process. This integrated method is based on coupling physical models and applying an optimization process. It is illustrated through the example of the layout of a half-bridge Mosfet power module exposed to electric and thermal constraints. So, physical models and software components have been developed, in order to define the multidisciplinary design process and then to perform the layout optimization using the NLPQL algorithm.It is shown that, the proposed approach allows automatic data exchange between physical models and optimal configurations are proposed with reduced design time and risk. Therefore, this integrated approach shows a great improvement in the design of PM and multidisciplinary products.