Water pollution in the river basins has become a pressing issue affecting the economic and social development of the coastal zones. Defining the variation trend of water quality in river basins is the prerequisite for the control and management of water pollution in these areas. In view of the monitoring of water quality and its development trend, taking the main stream of Huaihe River as the research object, we established the TM （thematic mapper） and OLI (operational land imager) water quality remote sensing monitoring models using multivariate statistical analysis method based on the measured water quality data obtained with the existing water quality monitoring method and the synchronization of Landsat data. According to the trend analysis results of Huaihe River water pollution from 2006 to 2017, the absolute error between the calculated values of the TM and OLI water quality remote sensing monitoring models and the measured values does not exceed 0.17, and the relative error does not exceed 7.11%. The comprehensive pollution index of the main stream of Huaihe River from 2006 to 2017 calculated by the models is consistent with the actual monitoring results, and the models can be used in the dynamic monitoring the of main stream water quality of Huaihe River. So it is feasible to establish a water quality monitoring model combining remote sensing image data with typical section monitoring data. The results can provide a new approach to the dynamic monitoring of water quality in large scale river basins and the protection of ecological environment.