The present specification discloses an image decoding method. The image decoding method according to the present invention comprises the steps of: determining a prediction mode of a current block to be a merge mode with motion vector difference (MMVD); deriving a merge candidate list for the current block; deriving a predicted motion vector for the current block by using the merge candidate list; deriving a residual motion vector for the current block; and deriving a motion vector for the current block by using the predicted motion vector and the residual motion vector, wherein only a part of candidates in the merge candidate list may be used to derive the predicted motion vector.