Based on the infrared spectra of Lophatheri Herba and Commelinae Herba, one-dimensional infrared spectra, second derivative spectra and two-dimensional correlated spectra were used to find out the differences between Lophatheri Herba and its imitations, respectively. The common peak ratio and variant peak ratio dual-indexes sequential were calculated and established according to infrared spectra of eleven batches of herbs. Infrared spectral data of Lophatheri Herba cluster analysis was applied to explore the similarity between each sample. The grouping results trend of sequential analysis of dual-indexes and cluster analysis was accordant. The results showed that the differences could be found by multi-level identification, and the source and the quality of the herbs could be effectively distinguished by the two analysis methods. Infrared spectroscopy, used in the present work exhibited some advantages on quick procedures, less sample required, and reliable results, which could provide a new method for the identification of traditional Chinese medicine with the imitations and adulterants, and the control of quality and origin.