A music shoe includes a shoe body for covering a person's foot, and a sound control unit installed in the shoe body. The sound control unit has a control circuit board with a voice database containing voice data, a speaker electrically connected to the control circuit board for voice output, and a plurality of buttons respectively electrically connected to the control circuit board and partially exposed to the outside of the shoe body for pressing by an external force to trigger the control circuit board to output a corresponding voice signal through the speaker.