This chapter deals with the integration of user's perceptions into the design process of products. Beyond the traditional approaches using predefined Design of Experiments (DOE) to study the effect of design variables of stimuli on human perceptions (classical in sensory analysis), methods based on Interactive Evolutionary computation (IEC) present a great interest. They are efficient to deal with complex optimization problems, where the derivative is unavailable, and can take into account perceptual assessments in the course of the solution search. In this chapter, we propose to show how Interactive Genetic Algorithms (IGA), a particular class of IEC, can be implemented for sound design. We illustrate the method with a particular application, the design of alert sounds for Electric Vehicles (EVs). EVs being very quite at low speed, it is necessary to add warning sounds but this can represent an annoyance if they are poorly designed. We describe a perceptual experiment in which the detectability and the unpleasantness of synthesized sounds are optimized with IGA. Results show that the quality of the sounds designed by the IGA method is significantly higher than different propositions of a designer, validating the relevance of the approach.