Fuel cell vehicle (FCV) is a sort of clean energy vehicle with zero emission. And its noise source layout spread out the front and rear part of the FCV. As a result, it is called distributed noise source. Because the interior noise of the FCV is composed of multiple noise sources, the linearity of the interior noise of the FCV is worse than that of the conventional cars. Therefore, the sound quality of the FCV should be improved so that the customers can enjoy the clean energy and comfortable environment simultaneously. Spectral characteristics of noises measured inside the FCV at idle are analyzed. The loudness value is used to evaluate the interior noise of the FCV. Based on evaluation results, the frequency ranges for improving the sound quality of interior noise are put forward. Combined with the spectral analysis results, the parametric filtering method is used to prepare 29 sound samples to find the critical frequency which affected the human ear’s subjective sensation seriously. A virtual validation about the effect of the design parameters on the improvement of the sound quality can be achieved by this method with advantage of time-saving and cost effective and without modifying the physical structure. Hence, the tendency of better solution will be found. To convert the sound quality problem identification into physical problem identification, the noise sources identification experiment is carried out by a single run method to quantify the noise sources which mainly contribute to the critical frequency of the interior noise spectrum. As a consequence, the main noise sources are identified as fuel cell auxiliary system.