Abstract: The sifted Lilium brownii sample powder (2 g) was taken, and 150 mL of 80% (volume fraction, the same below) ethanol solution was added. The mixture was treated by ultrasonic for 2 h and extracted by heating reflux for 2 h, and suction filtration was conducted while it was hot. The residue obtained was extracted again by heating reflux for 2 h with 150 mL of 80% ethanol solution. After suction filtration, the filtrates were combined, and concentrated by reducing pressure. The residue was dissolved by ultrasonic with 3 mL of water and 7 mL of methanol, and the mixed solution was centrifuged at speed of 12 000 r·min^-1 for 20 min. The supernatant was taken, and filtered through 0.45 μm microporous membrane. The filtrate was injected into high performance liquid chromatograph, and gradient eluted by mixed solutions composed of acetonitrile and 0.1% (volume fraction) phosphoric acid solution at different volume ratios as mobile phase. The Similarity Evaluation System for Chromatographic Fingerprint of Traditional Chinese Medicine (2004 A edition) was used for establishment of 11 batches of Lilium brownii samples from different origins, and the different markers that had great influence on the quality of Lilium brownii were screened by hierarchical cluster analysis (HCA), principal component analysis (PCA) and orthogonal partial least square-discriminant analysis (OPLS-DA). The 3 phenolic glyceride glycoside active components, including Regaloside H, Regaloside A and Regaloside B, were determined. As shown by the results, a total of 40 common peaks were identified by fingerprints of 11 batches of Lilium brownii samples from different origins, and the similarities were greater than 0.950. The results of HCA, PCA and OPLS-DA were consistent, and 11 batches of Lilium brownii samples were divided into 5 categories. Five different markers were screened, including Regaloside H, Regaloside A, Regaloside B and two unknown compounds. Linear relationships between the mass concentrations and peak areas of Regaloside H, Regaloside A and Regaloside B were kept in definite ranges, with detection limits (3S/N) in the range of 0.008 5-0.319 0 mg·L^-1. Test for recovery was made by standard addition method, giving results in the range of 98.0%-105%, and RSDs (n=6) of the determined values were in the range of 1.1%-2.2%.