Abstract: In the perioperative process, the patient's mean arterial pressure(MAP) value was used to reflect the anesthesia depth.As the mean arterial pressure was a time-varying and nonlinear variable, fuzzy RBF neural network control system in depth of anesthesia was studied.The detected MAP value was fuzzily processed and identified by fuzzy RBF neural so that PID controller tuned various parameters on-line by fuzzy RBF neural network.Fuzzy RBF neural network controller performance simulation was realized by software MATLAB, the results showed that the fuzzy RBF neural network control system for anesthesia depth control has better dynamic responses performance.