The multilayer video signal decoding method according to the present invention is characterised in that whether or not a corresponding picture in a lower layer is being used as an inter layer reference picture for a current picture in an upper layer is determined on the basis of a time level identifier for the lower layer, and if the corresponding picture in the lower layer is being used as an inter layer reference picture for the current picture in the upper layer, inter layer prediction of the current picture is carried out using the corresponding picture, and the current picture is restored using a prediction sample obtained by means of inter layer prediction and a residual sample of the current picture.