Abstract: A remote sensing monitoring model of sea surface temperature(SST) in Hebei coastal sea was build by correcting the Split window algorithm for retrieving SST with MODIS data. The retrieval results and corresponding actual measured results were compared at last. The results showed that the correlation coefficient of this model was 0.984, the total average error was 0.4℃, the average of relative error was 2.3%, the RMS error was 0.5℃.The interannual variability of SST in Hebei coastal sea was analyzed by the model between 2012 and 2014. The result shows that the model is preferably useful in quantitative retrieval SST in Hebei coastal sea.