The mass mortality of cultured scallops Chlamys farreri often occurs in summer and brings huge loss to farmers. However, the loss could be reduced greatly by transferring and renewing scallop cages or harvesting scallops before the occurrence of the massive death of scallops. A model for predicting the death of scallops by using the principle of Artificial Neural Network (ANN), the method of Back Propagation (BP) network and MATLAB software has been developed. The data to build the prediction model were acquired from a two year (2002 & 2003) investigation on temperature, salinity, pH, NH₃-N, NO₂-N of seawater and protein concentration, acid phosphatase activity, alkaline phosphatase activity, superioxide dismutase activity, catalase activity of scallops serum in Sanggou Bay, Rongcheng, Shandong Province. We debugged the model repeatedly by changing the key parameters, input layer node number, hidden layer node number, sample number and epoch number. After 300 times of studies and training, the sumsquared error of the prediction model decreased from 67.46 to 0.009 1. The model was tested, and the prediction accuracy was 87.5%. It is the first time that ANN was used in the aquaculture disease prediction. This model has many strong points, such as data adapting well, study momentarily and predicting accurately. The present study presents a new way for disease prediction and control of aquaculture animals.