The disclosed method of processing a digital broadcast signal comprises generating one RS frame or two RS frames according to the RS frame mode, forming the data group including the mobile service data and signaling data, forming and outputting a plurality of data packets including the mobile service data in the data group, wherein the 118 data packets and K data packets are outputted for the data group, the K being scalable between 0 and 38, interleaving the mobile service data included in the plurality of data packets, wherein the data group includes a scalable data region including data included in the K data packets, wherein the scalable data region carries data included in the first RS frame when the RS frame mode is the single frame mode, or carries data included in the third RS frame when the RS frame mode is the dual frame mode.