Chlorophyll-a is an important index of water quality for seawater, which can indicate the state of algae reproduction, further more it can predict the disaster of red tide by prediction model. The content of Chlorophyll-a of seawater is affected by many physical-chemical factors, this complex relationship among them is difficult to be described by ordinary mechanism expression. In this paper, we use Fuzzy BP model to describe this complex nonlinear system, and give a dynamic estimate to the output variables of the system. The PCA(Principal Component Analysis) method had been used to reduce the dimension of the sample data, simplify the complexity of the model system, this measure can make the model has a faster convergence rate and a relative low dimension. The experiment illustrates that fuzzy BP model based on PCA method can give the prediction for the content of chlorophyll-a in seawater to some degrees.