The rivers with the distinct gradient of water quality in the southern region of China were selected as a case study. The objective of this study was to develop the monitoring and evaluating technology of the water quality based on C-band polarimetric synthetic aperture radar (POLSAR). The random rough surface scattering model to describe the electromagnetic scattering characteristics of polluted water was briefly introduced. The potential effect of organic pollutants to the scattering model and backscattering coefficient were explored. The simultaneously obtained POLSAR data and the measured water quality indexes were analyzed. By comparing the POLSAR data and the water quality indexes, it could be observed that the chemical oxygen demand (COD) was in proportional to the ratio between HH and VV backscattering coefficients, which matched the analysis based on electromagnetic scattering theory. A fitting model was proposed to retrieve the chemical oxygen demand by ratio between HH and VV channel backscattering coefficients using least square method, with the fit coefficient of 0.90. In this study, the model using the ratio between HH and VV backscattering coefficients was established, which was mainly based on the analysis of experimental results, and was also supported by theoretical interpretation.