A dynamic control model for the secondary cooling of slab casting is presented to reduce the difference between the actual temperature and the goal surface temperature of slab. The model, which is based on the BP neural networks for forecasting the temperature and the fuzzy neural networks for dynamically controlling the water in the secondary cooling in the continuous casting, could timely adjust and allocate the water according to the speed and temperature of slab. A series of tests have been conducted based on inputs of the No. 2 slab caster in a steel plant. It has been shown that the model, which integrate the charateristics of water controlling problem in secondary cooling into the temperature status of slab during the cooling process, can control the water in secondary cooling efficiently and dynamically according to the situation of actual production.