Knowing the inherent optical properties (IOPs) of water bodies is useful for many water environment studies and applications. To derive the IOPs from remote sensing reflectance, a multiband quasi-analytical algorithm (QAA) was modified and validated for the highly turbid Poyang Lake in China. In order to supplement and expand the dynamic variation range of the measured water optical properties, a Hydrolight simulated dataset was generated to develop a regional QAA (QAA710) for this area. The QAA710 model was then validated with simulated data, simulated data with random noise, and in situ data. The results show that the effects of random noise (within ±20% ) of remote-sensing reflectance on the derived total absorption coefficients (a t ) and the particulate backscattering coefficients (b bp ) by the QAA710 model are insignificant (a band-averaged mean relative error of 4.1% and 12.0%, respectively). The validation of in situa t shows a 28.6% mean relative difference. The model process, modeling data, and validation data introduce uncertainties into the derived results. These analyses demonstrate that the QAA710 model, based on the characterization of local environments, performs well in retrieving Poyang Lake’s IOPs.