Abstract: Previous validation results showed that there was large bias in the operational sea surface temperature (SST) products from the visible and infrared radiometer (VIRR) onboard the Fengyun-3C (FY-3C) polar-orbiting meteorological satellite released by the national satellite meteorological center (NSMC). According to the setting of thermal infrared channels of FY-3C/VIRR, the regional algorithms to estimate daytime and nighttime SST in the seas around China were developed based on the non-linear SST (NLSST) and triple window NLSST (TNLSST). The clear-sky sea matchup data was generated from the bright temperature of satellite thermal infrared channels and in situ data, and the coefficients of NLSST and TNLSST were acquired by the regression fitting method. The regional SST algorithm was validated using the independent matchup data, the results showed that the bias and standard deviation (SD) were 0.082 ℃ and 0.633 ℃ in daytime, −0.007 ℃ and 0.557 ℃ in nighttime. Also, taking the optimum interpolation SST as reference value, the regional SST was compared with the operational products from the NSMC, the results demonstrated the bias and SD were decreased from 0.047 ℃ and 0.743 ℃ to 0.031 ℃ and 0.641 ℃ for daytime data, and from 0.184 ℃ and 0.708 ℃ to 0.034 ℃ and 0.556 ℃ for nighttime data, respectively.