Inertially stabilized platform (ISP) is widely used in aerial remote sensing for stabilizing imaging sensors to improve imagery quality. Due to the nonlinearity, time variation, and disturbances of the ISP system, conventional feedback control algorithms cannot achieve precise attitude control. To realize the control performance with fast dynamic response and high stabilization precision, a least mean square based active disturbance rejection control (LMS-ADRC) system is designed to reject multiple disturbances. In particular, the servo system model of the ISP is firstly established. Then, based on the model, the LMS-ADRC controller is designed, in which a least mean square based approach is proposed to estimate the control loop gain. The applicability of the proposed method is validated by a series of simulations and vehicle experiments, and significant improvements are observed in the presence of model uncertainties and external disturbances comparing to the traditional ADRC and PID controllers.