The concern about vehicle passengers' safety increases along the years, this can be concluded analyzing the constant evolution of requirements through standards, defining minimum structural rigidity of vehicles under different extreme loads. Otherwise, highly competitive automotive market demands lighter, economic and cheaper projects. The ambient concern is also in evidence. This diversity of objectives, many times contradictory, that an engineer must handle during a vehicle development, became an arduous task to relate all these concepts in a harmonious way. In this context, is presented in this paper a proposal of methodology for a structural optimization, applying Genetic Algorithm, of a bus rollover. The simplified finite elements model of the bus is constructed with beam elements and, in places with localized large plastic deformation, plastic hinges are added to the structure. Experimental and numerical methods are used to determine the plastic hinge properties. With this numerical model of the bus rollover, parameters are extracted to compose the different proposals of objective functions and, thus, diverse optimized projects are analyzed, aiming to find the optimum project for each study. Finally, is concluded that the optimization methodology proposal represents a useful tool and can be applied during projects' development, shortening the process and helping on important decisions.