In this study, the hybrid Cramer-Rao bound (CRB) is developed for target localization, to establish the sensitivity ofthe estimation mean-square error (MSE) to the level of phase synchronization mismatch in coherent Multiple-Input MultipleOutput(MIMO) radar systems with widely distributed antennas. The lower bound on the MSE is derived for the joint estimation of the vector of unknown parameters, consisting of the target location and the mismatch of the allegedly known system parameters, i.e. phase offsets at the radars. Synchronization errors are modeled as being random and Gaussian. A closed-form expression for the hybrid CRB is derived for the case of orthogonal waveforms. The bound on the target localization MSE is expressed as the sum of two terms, the first represents the CRB with no phase mismatch, and the second captures the mismatch effect. The latter is shown to depend on the phase error variance, the num ber of mismatched transmitting and receiving sensors and the system geometry. For a given phase synchronization error variance, this expression offers the means to analyze the achievable localization accuracy. Alternatively, for a predetermined localization MSE target value, the derived expression may be used to determine the necessary phase synchronization level in the distributed system.