The deterministic multiple objective optimization model was constructed based on the characters, including water shortage risk, economic benefit and fairness, existed in water resources management system. Besides, the Me measure multiple objective optimization model was proposed due to the uncertainties of water resources system. Moreover, the Me measure can be transformed into necessary measure, possibility measure and credibility measure according to different pessimisticoptimistic parameters. The two models were applied to the real case of interwater users water allocation of Ganzhou District, Linze County and Gaotai County. The results showed that the Me measure multiple objective optimization model was more practical for water resources management and had more robustness and provided more water allocation alternatives compared with the deterministic multiple objective optimization model. The results indicated that the total water allocation form high to low was possibility measure, credibility measure, and necessary measure. Besides, the structural water shortage risk had negative relationship with allocated water resources, while economic benefit and GINI were positively affected by allocated water resources. Therefore, the necessary measure had more advantages in minimizing the objectives, and the possibility measure was more applicable for maximizing the objectives, and the credibility measure had the comprehensive results. What’s more, the total water allocation, structural water shortage risk and economic benefit was lessened with the increase of measure levels, while the GINI coefficient was increased with the increase of measure levels. Therefore, the decision makers can choose best water allocation schemes with consideration of reasonable pessimisticoptimistic parameter and measure levels.