In CFD (Computational Fluid Dynamics) verification of vehicle aerodynamics, detailed velocity measurements are required. The conventional 2D-PIV (Two Dimensional Particle Image Velocimetry) needs at least twice the number of operations to measure the three components of velocity ( u,v,w ), thus it is difficult to set up precise measurement positions. Furthermore, there are some areas where measurements are rendered impossible due to the relative position of the object and the optical system. That is why the acquisition of detailed velocity data around a vehicle has not yet been attained. In this study, a detailed velocity measurement was conducted using a 3D-PIV measurement system. The measurement target was a quarter scale SAE standard vehicle model. The wind tunnel system which was also designed for a quarter scale car model was utilized. It consisted of a moving belt and a boundary suction system. In the experiments the moving belt was switched off while the boundary suction system was on. Mean velocity was calculated using the arithmetical average of 50 images of PIV data.Areas where measurements could not be made are those around the vicinity of a wall or those around curved surfaces. Missing data points bring about streamline discontinuity in flow visualization. Therefore, the authors have developed a technique to supplement missing data without losing measurement data reliability. This technique provides good data for the boundary conditions and calculates mass and momentum conservation equations. As a result, when error spots were small or limited, interpolated data was sufficient for evaluation.Then, the streamlines were compared with CFD results. The CFD code used the Partial Cells in Cartesian coordinate method. The results showed good correlation and confirmed the propriety of the CFD measurements.