The work aims to overcome the hysteresis, nonlinearity and time variability of the pulp concentration control system, so as to improve the pulp concentration control performance. A BP neural network PID control technique was proposed with respect to the problem of pulp concentration control. A 3-4-3 BP neural network structure was constructed. Based on that, a mathematical model of BP neural network PID control was built. The adaptive adjustment of PID parameters by BP neural network was made. The simulation results showed that, the BP neural network PID control had faster convergence speed, less overshoot, stronger anti-interference ability and better robustness than the traditional PID control. The control method realizes the self-adaptive control of pulp concentration and provides an effective and feasible control method for optimal control of pulp concentration.