In the paper authors identify the multi-criteria model of the significance of drainage pumping stations in Poland. The identified model will allow to classification of drainage pumping stations in terms of their significance in water management of the country. The Characteristic Objects method COMET, which is one of the methods of multi criteria decision analysis (MCDA), was used to build the model. In many scientific articles, issues related to the water management are frequently resolved by using MCDA methods. However, majority of MCDA methods are exposed to the rank reversal paradox, which has a negative effect with respect to the reliability of the final result. In contrast to other MCDA methods, the COMET method identifies the complete model in the space of the problem and not only discrete values for selected objects. The COMET approach identifies the expert model taking to account local feedbacks by using fuzzy logic mechanisms. Three parameters were used to build the expert model for drainage pumping stations, such as drainage efficiency, total output of aggregates and surface drainage. The result is 27 fuzzy rules that define the exact model. Subsequently, this model is used to assess the significance of 270 selected drainage pumping stations in the country. Then, the empirical method to verify the correctness of the results is also presented. On the basis of the results, the primary classification of drainage pumping stations is made in terms of their significance in water management of the country.