Rotor flow fields are highly complex and strongly unsteady, posing significant challenges for numerical simulation using traditional grids. While overset unstructured grids can effectively accommodate the rotor’s complex motion and geometric features, they often suffer from poor numerical stability and low computational efficiency. To overcome these limitations, this study develops a robust and efficient implicit numerical method for overset unstructured grids based on the exact Jacobian matrix. The proposed approach integrates a combined hole-cutting strategy based on signed distance fields and distances to physical surfaces with a contribution-cell search algorithm using a Bounding Volume Hierarchy (BVH), enabling fast and automated overset grid assembly. Numerical stability is further enhanced through a compact least-squares reconstruction, while the exact Jacobian matrix is computed via the chain rule to improve accuracy and reversibility, reducing errors in the implicit linear solver. An adaptive CFL scheme is also employed to further increase overall computational efficiency.Validation against representative test cases demonstrates that the proposed implicit method achieves substantial improvements in both convergence and computational efficiency compared with conventional approaches.