A novel method for voice conversion based on Viterbi algorithm is proposed in this paper.This method uses the matrix of transition probabilities of the target speaker’ frames to represent the timing information of the speech sequence

and then determines the most appropriate component of the GMM by utilizing the Viterbi algorithm for converting each frame of the source speech.It avoids the spectral discontinuities caused by losing the relationship between the adjacent speech frames

and alleviates the spectral smoothing due to the weighted averaging in the traditional GMM-based algorithms and then enhances the formant.Both objective and subjective evaluation’s results have demonstrated that the proposed method improves the performance of the conventional voice conversion system based on GMM.