It is indicated that atmospheric correction for optical remote sensing is absolutely necessary for quantitative remote sensing. The atmospheric correction methods can be grouped into four types, namely, relative correction based on image information, linear regression based on ground spectra, radiative transfer models and composite models. Their advantages and disadvantages as well as their adaptability or effective factors are analyzed in this paper. Some thoughts on the future development of atmospheric correction for optical remote sensing are also put forward in the end of this paper.