A high-performance liquid chromatographic (HPLC) method was established to analyze 36 Chaihu (Radix Bupleuri) samples collected from three species (Bupleurum chinense DC., B. scorzonerifolium Willd. and B. smithii Wolff.). Addition of trifluoroacetic acid into the mobile phase resulted in fingerprint chromatograms with stable baselines. There were thirty-two characteristic peaks in the standard fingerprint of B. chinense DC. Different recognition pattern methods, including similarity analysis (SA), hierarchical cluster analysis (HCA), principal component analysis (PCA) and partial least squares-discrimination analysis (PLS-DA) were utilized to analyze the 36 samples based on the contents of chemical constituents. Consistent results from SA, HCA and PCA analysis illustrated the rationalisation for why B. smithii Wolff. was not quoted in the Chinese Pharmacopoeia and classified samples were in agreement with their species. PLS-DA loading plots showed the chemical markers which had the most influences on the separation among different species. However, SA, HCA and PCA could not differentiate between wild and cultivated B. chinense DC. as well as between samples from different provinces. HPLC fingerprint in combination with chemometric techniques provided a very flexible and reliable method for homogeneity evaluation and quality assessment of traditional Chinese medicine.