Vector maps are distributed frequently among various departments and subscribers, and traditional single-stage watermarking technology cannot provide powerful multi-copyright protection of them. In this paper, a novel multi-stage watermarking algorithm is proposed for vector maps, aiming at robustness enhancement and maximum-supported watermark capacity improvement. A watermark generation method is designed based on QR code to shorten watermark length, and watermarks are embedded into polar coordinates of map vertices so as to improve the robustness of the proposed scheme. Another method of copyright information combination is raised to keep watermark length invariant along with the increasing of the total stages of watermark data, and the maximum-supported stages of watermarks can be thus improved without affecting the robustness of watermarking algorithm. Moreover, each stage of copyright information occupies a specified space within the final watermark data, and non-interference among various watermarks can be guaranteed and different copyright data can be efficiently distinguished after watermark extraction. Compared with the existing researches, the proposed scheme can increase effectively the size of maximum-supported watermark capacity, and it has strong robustness under common geometric and non-geometric attacks, which have been validated by both theoretical analysis and comprehensive experiments.