An interlayer video decoding method includes: reconstructing a color image and a depth image of a first layer in relation to coding information of the color image and the depth image of the first layer obtained from a bitstream; determining whether a prediction mode of a current block of a second layer image to be decoded is a view synthesis prediction mode that predicts the current block based on an image synthesized from a first layer image; determining a depth-based disparity vector indicating a depth-corresponding block of the first layer with respect to the current block, when the prediction mode is the view synthesis prediction mode; performing view synthesis prediction on the current block from the depth-corresponding block of the first layer indicated by the depth-based disparity vector; and reconstructing the current block by using a prediction block generated by prediction.