To study the fingerprint of the volatile oil of Atractylodes lancea (Thunb.) DC., and to offer the characteristic data for the quality evaluation, GC-MS analysis was performed for 17 samples of different areas used as Atractylodes lancea (Thunb.) DC. Nine kinds of same components were selected. TIC profiles were evaluated by "Computer Aided Similarity Calculation". The characteristic peaks in chromatograms were identified by comparing mass data with literatures. Hierarchical clustering analysis was performed by SPSS based on the relative peak area (RPA) of identified peak to atractydin in 17 samples. The mutual mode fingerprint plots of genuine Atractylodes lancea (Thunb.) DC. have been established, the matching of active components was characteristic that atractylon, hinesol, beta-eudesmol, atractydin as (0.89 - 1.12): (0.11 - 0.15) : (0.48 - 0.61) : 1. The difference of resemblance of not genuine samples with genuine samples was remarkable. Two categorizations were clustered. Atractylodes lancea (Thunb.) DC. from genuine and Tangshan and Nanshan, Jiangsu Prov. were in a group, while those from Anhui and Hubei Prov. were in another group. The characteristic fingerprint of genuine Atractylodes lancea (Thunb.) DC. combined with the study of the matching of active components for the quality control and the resemblance calculation of fingerprints and SPSS hierarchical clustering analysis provided a new analytic method for the quality evaluation and discrimination of Atractylodes lancea (Thunb.) DC.