Abstract: The study on spatial heterogeneity of soil organic matter (SOM) is significantly important to soil fertility evaluation, soil carbon pool estimation, soil resources utilization and protection. However, many factors, such as the spatial variation of relief, soil material and classes, few hyperspectral satellite images, restrict the regional SOM reversing with remote sensing (RS). Soil samples in typical blacksoil area of Heilongjiang province were collected, Landsat TM image was transformed and processed, and then the regional SOM predicting model was built with statistical methods. Results show that SOM content in blacksoil area is higher and larger than 2%, determining the dominant effect of soil organci matter; The spatial heterogeneity of SOM is significant, and soils are exposed for long time because of special farming and climate in blacksoil area, all these determine that the area is fit for SOM retrieving with RS; There are significant correlation between SOM and reflectance of six bands in VNIR spectral region, and the maximum R2 is -0.710 at Band 3 (0.63-0.69 μm), and the second at Band 4 (0.76-0.90 μm), which corresponds to the laboratory analysis result of hyperspectral reflectance; SOM exponential model based on Band 2-4 reflectance is the best on precision and stability, which can uncover the spatial content of SOM in blacksoil area; The results can provide theoretical and technical support for improving RS retrieving of soil physic-chemical parameters, evaluating soil quality and carbon pool.