The present invention relates to an audio processing method for hearing protection in an ambient environment, includes steps of: sampling digital audio signals decoded to obtain a plurality of amplitude values of the digital audio signals; obtaining a default gain value; calculating an actual audio energy of the digital audio signals sampled during a predetermined period time according to the amplitude values sampled in the predetermined period time and the default gain value; collecting ambient noises to calculate an ambient noise level; obtaining a reference audio energy according to the ambient noise level; comparing the actual audio energy with the reference audio energy; and generating a hearing protection signal if the actual audio energy reaches the reference audio energy, whereby protecting users' hearing. The present invention also provides a corresponding audio processing apparatus.