This paper examined the atmospheric correction of HJ-1 CCD data using an approach of look-up table created from 6S model to acquire land surface reflectance. The result shows that the image after the correction is much better than the original one, indicating that the attenuation of NDVI can be rectified to improve the application of HJ-1 CCD data for remote sensing of vegetation. Aerosol optical depth is required for the correction. In order to ensure the precision of the correction, the authors suggest that the required aerosol optical depth can be retrieved from the hyper-spectral data on the same platform as the HJ-1-A data. As for the HJ-1-B data, the aerosol optical depth can be estimated by using the contrast reduction method with consideration of the land surface BRDF.