Abstract: Based on the WorldView-3 data, the principal component analysis (PCA) is adopted to determine the principal components reflecting iron oxides, and Al(OH)-bearing and Fe, Mg(OH)-bearing minerals, by which the altered mineral information in Pobei area of Xinjiang is extracted to predict the metallogenic prospect area. The principal component eigenvector matrix of WorldView-3 data is analyzed by PCA, and USGS spectral library is used to analyze the absorption reflectance spectra characteristics of altered mineral associations, which both determine the principal components reflecting altered mineral information. Based on the determined principal components, the density slicing method are used, with remote sensing interpretation map, for the extraction of altered mineral information. With the extracted altered mineral information, the metallogenic prospect area is predicted by combining with the ore-bearing strata, quartz veins related with mineralization, and NE-trending fault structure controlling the distribution of orebodies in the area.