The recently developed quasi-analytical algo-rithm (QAA) is a promising algorithm for deriving inherentoptical properties from ocean color. Unlike the conventionalsemi-analytical algorithm, QAA does not need a prioriknowledge of the spectral shape of chlorophyll absorption.However, several empirical relations, which may not be uni-versally applicable and can result in low noise tolerance, areinvolved in QAA. In this study, the Bayesian inversion theoryis introduced to improve the performance of QAA. In theestimation of total absorption coefficient at the referencewavelength, instead of empirical algorithms used in the QAAthe Bayesian approach is employed in combination with anoptical model that uses separate parameters to account ex-plicitly for the contribution of molecular and particle scat-terings to remote sensing reflectance, a priori knowledgeproduced by the QAA, the Akaike’s Bayesian informationcriterion (ABIC) for choosing the optimal regularizationparameter, and genetic algorithms for global optimization.Coefficients at other wavelengths are then derived using anempirical estimate of particle backscattering spectral shape.When applied to a simulated dataset synthesized by IOCCG,the Bayesian algorithm outperforms QAA algorithm, espe-cially in higher chlorophyll concentration waters. The rootmean square errors (RMSEs) between the true and the de-rived a(440) and bb(440) are reduced from 0.918 and 0.039m–1 for QAA-555 to 0.367 and 0.023 m–1 for Bayes-555, 0.205and 0.007 m–1 for QAA-640 to 0.092 and 0.005 m–1 forBayes-640, and 0.207 and 0.007 m–1 for QAA-blending to0.096 and 0.005 m–1 for Bayes-blending. Results of noise sen-sitivity analysis show that the Bayesian algorithm is morerobust than QAA. The recently developed quasi-analytical algo-rithm (QAA) is a promising algorithm for deriving inherent optical properties from ocean color. QAA does not need a prioriknowledge of the spectral shape of chlorophyll absorption. However, several empirical relations , this may not be uni-versally applicable and can result in low noise tolerance, are in solved in QAA. In this study, the Bayesian inversion theoryis introduced to improve the performance of QAA. In theestimation of total absorption coefficient at the reference wave length, instead of empirical algorithms used in the QAAthe Bayesian approach is employed in combination with anoptical model that uses separate parameters to account ex-plicitly for the contribution of molecular and particle scat-terings to remote sensing reflectance, a priori knowledgeproduced by the QAA, the Akaike’s Bayesian information criterion ABIC) for choosing the optimal regularization parameter, and genetic algo rithms for global optimization. Coefficients at other sources are then derived using anempirical estimate of particle backscattering spectral shape. When applied to a simulated dataset synthesized by IOCCG, the Bayesian algorithm outperforms QAA algorithm, espe-cially in higher chlorophyll concentration waters. The rootmean square errors (RMSEs) between the true and the de rived a (440) and bb (440) are reduced from 0.918 and 0.039 m -1 for QAA-555 to 0.367 and 0.023 m -1 for Bayes-555, 0.205 and 0.007 m -1 for QAA-640 to 0.092 and 0.005 m -1 for Bayes-640, and 0.207 and 0.007 m -1 for QAA-blending to 0.096 and 0.005 m -1 for Bayes-blending. Results of noise sen-sitivity analysis show that the Bayesian algorithm is morerobust than QAA.