A strategy for using combined multi-wavelength fingerprints together with chemometrics for the prediction of antioxidant components in Turpiniae Folium is presented. Turpiniae Folium, classified as Turpinia arguta Seem. from the Staphyleaceae family, is widely used in southern China as a traditional Chinese medicine. Plant extracts have pronounced antioxidant activity and, in the present study, the antioxidant capacity of 29 different samples was evaluated using a 2,2-diphenyl-1-picryl-hydrazyl (DPPH) radical scavenging assay. Antioxidant activity was expressed as the concentration at which 50% of the DPPH radicals were scavenged (EC50). Based on the contour plot obtained by high-performance liquid chromatography using a diode array detector, four wavelengths (225, 254, 313 and 370nm) were selected to construct the combined fingerprints. The data were preprocessed using baseline correction by wavelet transform, data scaling, correlation optimized warping and standard normal variate processing to obtain more suitable data for chemometric analysis. A partial least squares regression model with four latent variables was then constructed based on EC50 values and combined fingerprints. The model possessed satisfactory predictive ability, with an explained variance of 73.06% for X variables, 94.58% for Y variables and a root mean square error for prediction of 0.6588. Combining the regression coefficients of the calibration model with qualitative information, seven compounds that were responsible for the antioxidant activity of Turpiniae Folium were identified.