In Volume II, optimal filtering and estimation is used for water quality modeling and prediction. Since complex water quality systems are generally nonlinear, nonlinear estimation and filtering techniques are emphasized. They generally can be classified as Kalman filters, minimum variance estimator, maximum liklihood (Bayessain) estimator, and maximum a posteriori estimator. Because of computational difficulties, almost all the nonlinear filters are approximations. These various filters are applied to a typical water quality problem. The advantages and disadvantages of the various filters are compared based on the computational results.