XML text classification model by RVM Classifier is proposed in this paper. Relevance vector machine is based on a Bayesian framework, compared with support vector machine,relevance vector machine has a comparable generalization performance and requires much fewer kernel functions. In order to solve multi-class problem, the multi-class classification formulation of relevance vector machine is presented. 240 XML texts with six XML text types are applied to study the effectiveness of RVM classifier for XML text classification. The comparison between the classification results of RVM classifier and the actual values is given, and the comparison between the classification results of SVM classifier and the actual values is given. The experimental results show that the classification ability for XML text of RVM classifier is stronger than SVM classifier.