This chapter deals with the state estimation with noise rejection in a piezoelectric cantilevered actuator and its state-feedback control. The noises which come from the sensor used, strain gage, are important and should be filtered. For that, we employ the classical Kalman filtering for their rejection and for the state estimation and we apply afterwards a state-feedback control with integral action to improve the general performances of the actuator. However, as the actuator exhibits hysteresis nonlinearity, we propose first its linearization thanks to a feedforward control before application of the above filtering and feedback control. The experimental results confirm the efficiency of the approach and demonstrate the interest of the method for precise positioning such as in micropositioning applications.