This paper used gray correlation analysis method to evaluate the water quality of Haihe River basin. However, the traditional grey correlation analysis has drawbacks due to subjectivity in calculating relatedness. To overcome this defect, this paper chosen two kinds of improved gray correlation analysis methods for the water quality evaluation, and compared the advantages and disadvantages of the two methods. The combined application of improved gray correlation analysis method and correspondence analysis method can reach to scientific and reasonable evaluation conclusions.Meanwhile, this paper predicted water quality using grey prediction model. The results showed that: (1) Combination of the improved AHP weights of gray relational analysis and correspondence analysis can provide a nice way to understand basin water quality and reach to a rational evaluation result; (2) The main sources of pollution in the Haihe River basin is agricultural non-point source pollution and sewage pollution.