Proposed are a method and an apparatus for encoding or decoding video, which: in a video encoding and decoding process, determine whether an intra-prediction mode of a current block is a DC mode; if the intra-prediction mode of the current block is the DC mode, when the height and the width of the current block are identical, determine, as a DC value, an average of upper reference samples of the current block and left reference samples of the current block; when the height and the width of the current block are different, determine, as an upper DC value, an average of the upper reference samples of the current block, determine, as a left DC value, an average of the left reference samples of the current block, and determine, as a DC value, a weight average value determined by combining predetermined weights with the upper DC value and the left DC value; and perform intra-prediction of the current block on the basis of the DC value.