Halfshafts are very important components from vehicle powertrain. They are the element responsible to transmit torque and rotation from transmission to wheels. Its most basic design consist of a solid bar with joints at each extreme, however, Depending of its length, the natural frequency of first bending mode might have a modal alignment with engine second order, resulting in undesired noise on vehicle interior. Many design alternatives are available to overpass this particular situation, like adding dampers, use tube shafts or use link-shafts, however, all of them are cost affected. The aim of this study is to propose an optimal halfshaft profile for a solid shaft to be machined from a rough bar, pursuing the lowest frequency the possible for the first bending mode by changing its diameter and, as a consequence, the mass and bending stiffness at appropriate regions. The study is divided in three main stages: initially, a modal analysis of a halfshaft is done a vehicle to determinate its dynamic behavior when assembled on vehicle. Second, a CAE model is generated to reproduce results from modal analysis and validate the boundary conditions. On the last stage, the CAE model is submitted to an optimization routine through genetic algorithms, varying the diameter at different sections of the shaft and determining its natural frequency as a function of its profile. Once an optimal design has been achieved, a prototype has been made to validate the results. It has demonstrated that the proposal had an impact on the natural frequency, however, further investigation is recommended to extend the potential of the method.