In the view point of driving safety, the crosswind sensitivity of a vehicle becomes more important, as the driving speed in highway gets higher in these days. The sensitivity of a vehicle to crosswind depends on many factors, including the design of the suspension and aerodynamics of the body, etc. However, the knowledge about this phenomenon has still to be improved, in order to develop vehicle with optimum characteristics for crosswind stability.In this research, the physics behind the sensitivity of a vehicle is discussed in detail through various kinds of virtual test using computer aided engineering (CAE) simulation scheme. In the first, a reliable simulation model for vehicle, driver, wind generator and interactions among them is built. This simulation model is verified by comparison with test results of real vehicle. Then, the sensitivity analysis is carried out to find out the most influential design parameters. Finally, an automated design optimization process is carried out based on sensitivity analysis results by connecting virtual test environment to process integration & design optimization (PIDO) tool.With this procedure, both deep understanding of interaction among vehicle, driver and crosswind, and optimized set of design parameter to make a vehicle robust against the disturbance by crosswind can be achieved.