Electrohydraulic servo system has been widely utilized in various industries. However, robust adaptive control logic is quite necessary for the electrohydraulic servo to be used in wide operation range because of its complex and nonlinear characteristics with time-varying parameters. This study developed a self-turning regulator (STR) as a nonlinear adaptive control logic robust to noise and disturbance. Properties of electrohydraulic valve was analyzed firstly, and its nonlinearity was compensated by an inverse nonlinear model. The entire dynamic model equation was obtained by system identification techniques, namely, online forgetting factor recursive least square (FFRLS) algorithm. Then robustness of adaptive control scheme in this study against noise was also simulated. The results show that robustness to noise of the adaptive control logic works well and STR control algorithm can effectively compensate system nonlinear characteristics.