Abstract: Though the support vector machine has been successfully applied in hyperspectral imagery classification, it has also several limitations. Relevance vector machine is a sparse model in the Bayesian framework, its mathematics model doesn’t have regularization coefficient and its kernel functions don’t need to satisfy Mercer's condition. RVM presents the good generalization performance, and its predictions are probabilistic. In this paper, we firstly analysis the disadvantages of the support vector machine for hyperspectral imagery classification, and then a hyperspectral imagery classification method based on the relevance machine is brought forward. We introduce the sparse Bayesian classification model, regard the RVM learning as the maximization of marginal likelihood, and select the fast sequential sparse Bayesian learning algorithm. Through the experiments of OMIS and AVIRIS images, the advantages of the relevance machine used in hyperspectral remote sensing image classification are given out.