A pop sound suppression method, an audio output circuit, and a terminal suppress a pop sound that is generated when an audio output circuit is in an alternating current (AC) coupling structure. The output circuit includes an output power amplifier, a common-mode voltage buffer, a reference voltage generation circuit, a powered-on pop sound suppression switch, and a common-mode switch. The powered-on pop sound suppression switch is configured to control, in a power-on process of the audio output circuit, a voltage level of an output node to be zero. The common-mode switch is configured to control, when a reference voltage level of the reference voltage generation circuit is zero, the voltage level of the output node to be equal to the reference level.