A feedback linearization controller for hydraulic system based on error pretreatment was proposed to weaken the influence of the strong nonlinear characteristic and parameter timevarying characteristic, in which RBF neural network was adopted to identify the hydraulic system and establish mathematical model. In order to obtain the position output with fast and high precision, the error was pretreated to adjust the error gain, which made the gain vary with the error. When the error was reduced, the gain was increased to accelerate the response speed and improve the control accuracy; when the error was increased, the gain was reduced to prevent the system overshoot. Based on the error pretreatment, feedback linearization method was used to design the controller and adaptive law. Simulation results show that the response speed and control accuracy of the system under error pretreatmentbased feedback linearization control are better than those of feedback linearization control.