A new Chinese chunking algorithm is proposed based on Naive Bayes model and semantic features. Through the analysis of Chinese chunking task, Naive Bayes model that combines different types of features were applied for its rapid performance of training and test. Semantic features were utilized to further improve the accuracy. Experimental results on the Chinese chunking corpus of Chinese Penn Treebank show that the algorithm achieves impressive accuracy of 92.8% in terms of the F-score.