An improved method for recording audio and video information together on a video disk or the like wherein the position of the audio information is made to coincide closely with that of the video information. Both the video and the audio information are divided into blocks. At the start of recording of one block of video data, recording of a block of audio data one block behind is commenced. At the end of the recording of the block of audio data, if the time remaining for recording the block of video data is less than a predetermined value, a subsequent block of audio information is recorded, while if the time remaining is greater, a block of inserted data, such as audio silence, is recorded.