An asymptotically efficient estimator for determining the position and velocity of a moving target from time delay and Doppler shift measurements in the presence of location uncertainties in multiple-input multiple-output (MIMO) radar systems with widely separated antennas is presented. It is considered that the actual positions of transmitters and receivers are not available. By taking both the location and measurement errors into account, a novel closed-form two-stage weighted least-squares solution to the problem is developed and analyzed. In order to generate a substantial improvement in the performance of the method, a weighting matrix is employed at each stage of the process. The accuracy properties of the method as well as the Cramer–Rao lower bound (CRLB) for target localization accuracy are derived in the case of Gaussian observations. The proposed algorithm is shown analytically to achieve the CRLB under small noise conditions. Simulations are included to examine the algorithms performance and corroborate the theoretical developments.