AbstractIn the 1990s, the emergence of MODIS has caused to try new something in soil. This paper explores the soil extraction of secondary classification and accuracy based on computer supervised classification with MOIDIS data, topographic map and soil map data in fuyu country, heilongjiang province. The research indicates it's effective for using MODIS to extract soil speies classification and the extraction of homogeneous dividation can be realized and cultivates high accuracy, which indicates that MODIS spectral differences on soil target faint dividation have certain inversion ability and soil information extraction can be attained certain value of the classification results based on the spectral characteristics and auxiliary knowledge. It is a useful supplement for black soils secondary classification based on MODIS data and it is an effective means for black soil monitoring and protection.