The impact of multi-block mesh refinement on automobile aerodynamics was investigated using three refinement methods: (1) overall mesh refinement, (2) ground mesh refinement, and (3) prism layer control. The open source CFD code, OpenFOAM V10, was utilized and simulations were conducted on the full-scale Notchback DrivAer model. For each method, the effects of mesh refinement were analyzed by refining different regions. In the overall mesh refinement study, successive multi-block grid refinement regions were established - the grid size becomes finer as the block is closer to the car surface. The lift coefficient was affected by grid refinement in regions near the automobile body. The ground mesh refinement study was conducted by controlling two grid sizes related to the ground region: volume mesh and surface mesh. The prediction of drag coefficient was fairly robust with ground region refinement, and the lift coefficient was more accurate with a smaller ground region grid. However, an excessively small surface grid reduced the accuracy of both drag and lift coefficients. In the prism layer study, the accuracy of the drag coefficient was independent of the prism layer setup, while the layer thickness and expansion ratio had a significant impact on the lift coefficient.