An algorithm for audio watermarking based on multiresolution analysis is proposed.The algorithm deals with the audio signal using a wavelet decomposition to get the time-variant low frequency components of the audio signal.Then,the spectrum of the low frequency components is spread and the watermark is embedded into the low frequency components spread by using the MDCT(modified discrete cosine transform).The better veracity of extracting the watermark is assured by using error correction.By using the method,the amount of the embedded watermark bits is increased and the security and imperceptibility of the embedded watermark are ensured.In the mean time,the watermarking scheme is also robust to the common signal processing.