A fuzzy optimization model is developed for the seasonalwater quality management of river systems. The model addresses the uncertaintyin a water quality system in a fuzzy probability framework. The occurrenceof low water quality is treated as a fuzzy event. Randomness associated withthe water quality indicator is linked to this fuzzy event using the conceptof probability of a fuzzy event. In most water quality management models therisk level for violation of a water quality standard is constrained by a preassigned value through a chance constraint. In the fuzzy risk approach a range of risk levels is specified by considering a fuzzy set of low risk, instead of using a chance constraint. Thus the two levels of uncertainty, one associated withlow water quality and the other with low risk, are quantified and incorporatedin the management model. The model takes into account the seasonal variationsof river flow to specify seasonal fraction removal levels for the pollutants.Fuzzy sets of low water quality are considered in each season. The membershipfunctions of these fuzzy sets represent the degree of low water quality associatedwith the discrete states of water quality in a season. The fuzzy risk of lowwater quality in a season at a checkpoint in the river system is expressedin terms of the degree of low water quality and the steady state probabilitiesassociated with discrete river flows. Considering the fuzzy goals of the pollutioncontrol agency and dischargers, and the crisp constraints, the water qualitymanagement problem is formulated as a fuzzy optimization model. The modelsolution gives seasonal fraction removal levels for the pollutants. Applicationof the fuzzy optimization model is illustrated with a hypothetical river system.