Abstract: The traditional sampling methods of biomass measurement could be used for local data up scaling, but these methods usually imported errors and could’t acquire continuous spatial information. In this paper, the China’s self-developed medium resolution HJ-1A/1B satellite data was processed and analyzed for ginkgo biomass estimation in Taixing, Jiangsu province. First, the Ginkgo biloba stands were extracted by vegetation sensible bands and eigen components. Second, by using the field estimated volume and the eigen components, the multivariate regression models were selected to estimate volume after the comparison of the statistical response co-efficiencies of both the single factor regression and the multivariate regression models(the R² of multivariate regression models is 0.72; the RMSE is 0.18). Finally, the total amount of biomass(745 144.14 t)was estimated by local BEF(biomass expansion factor), as well as the total amount of carbon storage and carbon density mapping. The results indicated that the total volume of Ginkgo biloba in 2010 had achieved the goal of the anticipated amount, but the carbon density was relatively low and the spatial distribution was inhomogeneous. So, in order to explore the carbon sink potential and increase the carbon sequestration capacity, the local efforts should be placed in increasing the planting area and strengthening young-and middle-age forests tending, as well as popularizing Ginkgo biloba agroforestry model etc.