Atmospheric correction is essential in satellite data processing to reduce atmospheric and lighting effects by studying the physical parameter of the earth's surface. In this study, ATCOR and 6S algorithms were evaluated for Sentinel-2 over paddy field area. In this evaluation, level-1C Top of Atmosphere (TOA) Sentinel-2 image was used as an input data. The spectral pattern analysis of the results was used to assess the reliability of the methods. As a result, the both methods produced the correct spectral patterns. Moreover, the results showed that the longer the wavelength the less the improvement values. It starts from the blue band, 43, 2% for the fallow phase and 60% for the vegetative phase of BOA corrected image. In the other hand, the NDVI values of fallow and vegetative phases that derived from the two methods are not greatly different.