In this paper, we propose the design of artificial underwater retro-reflective markers and their recognition algorithm. Retro-reflective materials, which are used for some portion of the marker, are used to overcome the limitation of underwater visibility. Owing to the distortion in the image captured by an underwater camera, the recognition algorithm consists of image preprocessing and 6-DOF pose estimation. The recognition performance and pose estimation of underwater markers are tested to verify the feasibility of the proposed retro-reflective markers in the real sea environment.