A large area outbreak of green tide poses serious negative impact on marine environment, fishery economy, and coastal tourism. Remote sensing technology has an advantage in macroscopic and dynamic monitoring of green tide in time, and is of significance for green tide treatment and prevention in time to reduce economic lost. We extract spectral reflectance characteristics of green tide from remote sensing data in coastal waters, and analyze spectral differences between green tide and normal waters. Towards the domestic satellite data of HJ-1 and GF-1, we develop a multispectral green tide index(MGTI)-multiband difference coupling algorithm that can be used to effectively detect green tide based on remote sensing technology. Meanwhile, remote sensing monitoring of green tide area by using Landsat7-ETM+ data is collected as reference, and also two conventional remote sensing algorithms, including normalized difference vegetation index(NDVI) and enhanced vegetation index(EVI), can be used for comparing with our developed algorithm. The obtained results demonstrate that the developed algorithm can be effectively applied into green tide monitoring in China coastal waters, and the monitoring accuracy is approximately 94%. This research can provide a theoretical basis and strategy for using domestic satellite data to monitor green tide in coastal waters.