Based on the fishing data including months, the number of vessels, the catch per day of the red flying squid (Ommastrephes bartramii) in jigging fishery by Chinese fishing fleet, and the data of the corresponding oceanic environment, i.e. longitude, latitude, the water temperature at 5 m (T005), 46 m(T046), 112 m(T112), 317 m(T317) under surface, chlorophyll a(CHA), and sea surface height anomaly(SSHA) in the North Pacific Ocean during June and November in 1998 to 2004, the BP neural networks model was applied to predict the emergence and distribution of fishing grounds of red flying squid after the standardization of CPUE. The total of 13 models with different hidden layers have been tested. The optimum model was selected by comparing the values of the simulation residual of each model structure statistically. The result shows that the BP neural networks model with 9 7 1 networks structure with 0.008 570 simulation residual only can be used for better predicting the fishing ground of the red flying squid in North Pacific Ocean.