A fast identification method of eleven genera of Chinese herbs in geraniaceae was developed by the combination of Fourier transform infrared spectroscopy with clustering analysis. FTIR spectroscopy was employed to identify and analyze eleven genera of Chinese herbs in geraniaceae. On the basis of a principal component analysis (PCA) model, three genera of Chinese herbs were rapidly classified by using the method of SIMCA clustering analysis. These samples could be successfully classified by SIMCA. Recognition rate and rejection rate reached up to 98%. The accuracy of clustering reached up to 91% during blind sample testing. It is concluded that in combination with clustering analysis, FTIR method provides an effective way to rapidly evaluate Chinese herbs in Geraniaceae.