Most water environmental factors have a nonlinear correlation. To assess sea water quality, the Back Propagation (BP) artificial neural network model was established based on the concept and principle of artificial neural network. Taking DO, COD, DIN, PO43- and petroleum as the evaluation factors, the BP neural network was established to evaluate the water quality for Dagu River wetland. The results indicated that the sea water quality of Dagu River was worse than the class Ⅳ water quality standard, and the water quality of its surrounding area were or exceeded the class III water quality standard. The survey sea area was mainly affected by inorganic nitrogen and phosphate. The study found that the number of inorganic nitrogen content that exceeded Ⅳ levels accounted for 59.3% of all respondents, and that eutrophication was obvious. These results indicated that BP neural network method was reasonable in design and higher in generalization compared with single factor evaluation and Nemerow pollution index, and that it is an objective, effective and practical environmental quality evaluation method. Thus, BP artificial neural network was a better level of fast, handy and valid ability to evaluate the sea water quality.