Since the voice conversion (VC) technology becomes widely used, there have been a lot of methods proposed by researchers in recent years. In this paper, we introduce the various applications and four kinds of performance evaluation approaches for VC. Furthermore, we also give an overall review of VC kernel methods and conclude with some viewpoints. This study involves four popular kernel methods, which are vector quantization, Gaussian mixture model, artificial neural network, and hidden Markov model. The principle and the property of each method are elaborated in this paper.