The level assessment of water quality has great significance for the prevention of water pollution; Its assessment is influenced by many factors. To assess the level of water quality in the Ningxia section of the Yellow River accurately, the intuitionistic fuzzy sets -TOPSIS model is introduced at first; secondly, the decisive matrix about the intuitionistic fuzzy sets is established, and the index weight coefficients considering the uncertainty of assessment indices are determined by using the Entropy weight method, then the weighed decisive matrix is obtained. Finally, the degree of membership at different levels of the water level is determined. Based on the ranking sequence of degree of membership, the water quality level corresponding to the maximum degree of membership is the final assessment level. The conclusions are drawn that the accuracy of level estimation of water quality in the Ningxia section of the Yellow River is very high using the intuitionistic fuzzy sets model in comparison with the current specifications, and the method is feasible for the level assessment of water quality, so it provides a new method and thoughts to assess the level of water quality in the future.