The H.264 video coding standard, owing to its excellent compression efficiency, has become the key technology of many video applications. But the H.264/AVC video is very vulnerable to data errors or packet losses. In this paper, we propose a new H.264 error concealment technique that exploits the depth relevance between spatially or temporally neighboring blocks. The motion vectors gathered from received H.264 slices are used to estimate the depth values and construct the depth maps. New motion vectors derived from depth maps are then added to the set of candidate motion vectors for concealing a lost block. The best motion vector is determined by an external boundary matching algorithm. Experimental results show that the proposed method is effective in improving the quality of received video without explicitly generating and transmitting the depth maps. Compared to the conventional method without incorporating the depth relevance, the proposed method increases the PSNR by 0.5 to 3.6 dBs.