Abstract: A method and apparatus for automatically compensating a sound volume are disclosed. The method includes the steps of determining whether a channel conversion key and an audio input switching key are inputted, producing an energy of an audio signal when the channel conversion key is inputted, and comparing an energy value of the audio signal detected for a predetermined time period with an energy value obtained from a previous step thereof and compensating for as much volume as a difference obtained by the comparison. The method together with the apparatus considers a symmetrical characteristic of an audio signal, calculates a voice energy with regard to a half signal of the audio signal, converts the calculated signal to a simplified value, and automatically compensates for the volume accordingly, thereby removing redundant manual steps.