Abstract: AimTo establish a high performance liquid chromatographic fingerprint for the quality control of rhizoma Chuanxiong, a traditional Chinese medicine derived from the root of Ligusticum chuanxiong Hort.. MethodsAn on-line optimized HPLC-DAD-MS technique was employed. The HPLC analysis was performed on a Waters Symmetry C₁₈ column (150 mm×4.6 mm ID, 5 μm) with a Waters Spherisorb S5 ODS2 (10 mm×4.6 mm) guard column. The mobile phase consisted of A (methanol) and B (0.25% acetic acid). Components were separated using the following gradient profile: 32% B at 0-3 min, 32%-85% B at 3-33 min, 85%-100% B at 33-52 min; flow rate was 0.7 mL·min^-1. DAD was set from 190 to 400 nm, the fingerprint was monitored at 294 nm. All mass spectra were acquired in the positive ion mode with electrospray ionization; the full scan mass spectrum was recorded over the range ofm/z100-800. Nine samples from three companies were analyzed; the main characteristic peaks were identified based on the comparison of UV and MS spectra of each analyte with that of authentic compounds and literature data. ResultsThe HPLC fingerprint was established based on the analysis of nine rhizoma Chuanxiong herbal samples supplied by three companies. Twenty-one characteristic peaks were found in all nine samples. These peaks were classified into four groups: group I at 0-12 min, three peaks were found, and the marker peak 3 was confirmed as ferulic acid; group II at 12-24 min, four peaks were found, and the marker peaks 4 and 5 were identified as senkyunolide I and senkyunolide H; group III at 24-32 min, there were seven peaks, and the marker peaks 9, 11, 13 and 14 were elucidated as senkyunolide A, coniferylferulate, ligustilide and 3-butylidenephthalide, respectively; group IV at 32-50 min, seven peaks were observed, and the marker peaks 15 and 17 were identified as riligustilide and levistolide A. The peak areas of 13 main peaks with normalized peak area (1% were determined. Using the most abundant peak 13 as the reference peak, the calculated relative retention times (T^R of the characteristic peak/ T^R of the reference peak) among nine samples were consistent (RSD≤1%), while the calculated relative peak areas (peak area of the characteristic peak/peak area of the reference peak) among nine samples were significantly different (p<0.001), indicating that all nine samples tested contain similar 13 main components with different quantities. ConclusionThe established HPLC fingerprint is very specific, and can be used to evaluate the quality consistency of different rhizoma Chuanxiong herbs.