Excessive wind noise is one of the most complained problems by owners of new vehicles as evidenced by JD Power Initial Quality Study (IQS) in recent years. After the vehicle speed surpasses 100 km/h, wind noise is gradually becoming the dominant noise source. In an effort to reduce aeroacoustic noise level, Beamforming (BF) is a very effective noise source identification technique used during vehicle wind noise development phases. In this work, based on the planar BF methodology, a large semi-circle microphone array is designed in accordance with the desired resolution and dynamic range pertaining to actual noise source distribution on a typical passenger vehicle. Acoustic array calibration and mapping deformation correction are accomplished by multi-point source method, and the Doppler Effect due to wind is corrected by the location calibration method. The established system is applied to identifying the exterior noise sources on the driver-side of a passenger vehicle in the wind tunnel. The noise sources induced aerodynamically in the area of front wheel housing, rear wheel housing and side mirror are separated successfully. Furthermore, the noise source strength from three different side mirrors is ranked, which correlates well to the interior loudness ranking measured by artificial heads.