In vehicle aerodynamics analysis considering actual traveling conditions, it is known that the flow around the rotating wheel interferes with the wake of the wheel arch, the flow ejected from the engine room, and the flow under the floor. This significantly affects the aerodynamic drag value. In particular, the fuel consumption measurement method by the World Wide Harmonized Light Vehicle Test Procedure has been carried out since FY 2018 in Japan. A calculation condition on computational fluid dynamics is required to be accompanied with wheel rotation. However, it is not easy to carry out simulations with wheel rotation for a full-vehicle model using a general-purpose method with an unstructured grid or voxel lattice. This method has difficulty in performing calculations faster and easier. On the other hand, the authors proposed a method for complex geometries using a hierarchical Cartesian grid and a topology-independent immersed boundary method, making it possible to handle “dirty” CAD data directly. Several research results for full-vehicle aerodynamics have been reported with high parallel computing efficiency and high practicality. In this study, this method is applied to vehicle aerodynamics analysis with wheel rotation, and is compared with experimental values measured in a wind tunnel with a moving belt facility. Full-vehicle aerodynamics analysis was performed on six configurations, and the difference in the aerodynamics forces was predicted. The grid resolution was about 6 mm, the number of computational grids was about 120 million, and the calculation time for an unsteady flow of 2 s took about 15 h in stationary cases and about 39 h in full-rotation cases. As a result, the drag value could be predicted within a range of ΔCdA ± 0.015. In addition, a difference in the flow field around the rotating wheel was observed owing to the difference in the numerical configuration of the rotation condition.