In this study, an inexact two-stage stochastic credibility constrained programming (ITSCCP) method is developed and applied to the water quality management system. It incorporates the credibility constrained programming, two-stage stochastic programming and interval programming within an optimization framework, so that uncertainties expressed as intervals, probabilistic and possibilistic distributions can be effectively communicated. The study area is the Xiangxi River Basin of the Three Gorges Reservoir (TGR), the largest reservoir in China. The significant advantage of the ITSCCP water quality management model is that it can reflect the tradeoffs between the predefined economic targets and the associated environmental penalties, as well as the fuzziness of the pollution load capacities. Stable interval solutions can be obtained by using the two-step interactive solution algorithm. A spectrum of potential pollution abatement options with varied levels of system-failure risk can be generated. The resulting solutions show that the proposed method can provide scientific bases for water quality management and other environmental systems problems.