We propose a novel content-aware three dimensional (3-D) image (two dimensional (2-D)-plus-depth) retargeting method using a 3-D saliency-based energy function that effectively describes perceptual importance of objects in 3-D space. The 2-D and depth images are retargeted together by carving out collinear seam paths corresponding to the one having the least 3-D saliency energy. However, resizing individual objects in different ratios without depth modification can cause surface shape distortion of the 3-D objects. To minimize such distortion, depth values for the objects are adaptively remapped examining depth histogram change. Experimental results indicate that the proposed method can preserve structures of important objects in 3-D space much more effectively than conventional methods.