In this paper, a sliding mode observer (SMO)-based model reference adaptive system (MRAS) estimator for sensor-less vector control of rotor-tied doubly-fed induction generators (RDFIGs) in wind turbine applications is presented. The idea behind the association of the SMO and MRAS in the proposed estimator is motivated by the improvement that the use of the SMO brings. The improvement includes a better dynamic response and robustness against parameter variations. The performance of the proposed SMO-based MRAS estimator is investigated through simulations and compared to the standard MRAS estimator under various operating conditions of the RDFIG.