The reservoir operator is in a critical position when the objectives are conflicting in nature, like releases to irrigation canals which reduces the reservoir storage while for getting more hydro power generation the storage should be high. This necessitates the study of trade-off analysis between the conflicting objectives and determination of compromised solution to get maximum benefits from the conflicting objectives. Therefore, when goals are uncertain & conflicting, fuzzy optimization is necessary to find the best compromised solution. In the present study, best compromised reservoir operating policies are developed using Multi Objective Fuzzy Genetic Algorithm - Non Linear Programming (MOFGA-NLP) hybrid model. This can be done in three steps. In the first step, models is solved by GA-NLP hybrid approach considering one objective at a time and determined the best and worst values of each objective function. Objectives are fuzzified by considering suitable membership function and the model is reformulated to maximize the level of satisfaction (λ) in step 2. In the last step, the reformulated model is solved and determined best compromised policy. The above model is applied to Nagarjuna Sagar reservoir located on river Krishna, India. Objectives considered in the present study are, maximizing the net benefits from irrigation & hydro power generation. The compromised reservoir operating policies are found for various reliable inflows (75%, 80%, 85% and 90%) entering into the reservoir by considering objective functions as fuzzy. The level of satisfaction is more for hyperbolic membership function compared to linear membership function.