In order to improve the control performance and trajectory tracking effect of hydraulic quadruped robot, the self-tuning fuzzy-PID control algorithm was applied to drive the hydraulic cylinders of leg joints, and the increment values of PID parameters were adjusted in real time. The mathematical model of valve-controlled asymmetrical hydraulic cylinder system was established, and the control problem caused by the nonlinear phenomenon and time-variance of plant parameters when the hydraulic cylinder was moving forward and reverse were analyzed. The control simulation was performed based on the co-simulation environment of AMESim and Simulink, and a physical prototype testing was done on the single leg experiment platform. Simulations and test results indicate that the self-tuning fuzzy-PID control algorithm shows the better control effects than conventional PID control in many respects, such as shortening the adjusting time, and restraining the impulse interference. This control algorithm can improve the dynamic tracking performance of robot legs, is easily applied in the projects and is helpful for hydraulic robot’s control.