Tongue inspection is an important diagnostic method in traditional Chinese medicine (TCM). However, due to its qualitative, subjective and experience-based nature, traditional tongue inspection has found very limited application in modem clinical medicine. In this paper, a novel system for tongue inspection based on hyperspectral imaging system is developed. The system includes image acquisition, feature extraction and classification functions. In this system, two kinds of quantitative features, spatial and spectral, are extracted from hyperspectral tongue images by using popular digital image processing techniques. Then, Bayesian networks are employed to model the relationship between these quantitative features and diseases. The preliminary results show that the system is sensitive to the abnormal tongues.