A method of predictive-decoding a multi-view image includes receiving a bitstream comprising information indicating whether each of pictures included in multi-view image sequences is a reference picture referred to by others of the pictures included in the multi-view image sequences for inter-view prediction and data about the multi-view image sequences; extracting the information indicating whether each of pictures included in multi-view image sequences is the reference picture and the data about the multi-view image sequences from the received bitstream; determining whether pictures included in the data about the multi-view image sequences are to be used for inter-view prediction decoding based on the extracted information indicating whether each of pictures included in multi-view image sequences is the reference picture; and performing predictive-encoding based on a result of the determination, wherein the information indicating whether each of pictures included in multi-view image sequences is the reference picture is included in a parameter for each of the pictures.