Abstract: The coral reef bleaching phenomenon has been frequently observed worldwide in recent years due to global climate change and extreme weather conditions. The applicability of remote sensing SST in coral bleaching early warning in the Beibu Gulf and coastal areas of Hainan Island is not known. In this study, we compared three commonly used remote sensing SST data (OISST, GHRSST and CoralTemp) and their applicability in coral bleaching early warning during summer. It showed that, GHRSST had the highest SST, CoralTemp the second and OISST the lowest, the bias of OISST was more obvious before 2013, and GHRSST and CoralTemp were close to each other after 2003; compared with in situ observations of buoys, from 2006 to 2020, the mean deviation and root mean square error were 0.03 ℃ and 0.92 ℃ for CoralTemp, 0.08 ℃ and 0.96 ℃ for GHRSST, and -0.25 ℃ and 1.21 ℃ for OISST, respectively, CoralTemp was more accurate in the study area; DHW calculated by the three remote sensing SSTs followed the same trend as those in situ SSTs calculated, with CoralTemp's results more closer to in situ SST. Therefore, among the three remote sensing SST data, CoralTemp is relatively more applicable in the study area. The findings of this paper will provide a reference for the research and application of coral bleaching early warning models in the study area.