Recently, multidimensional scaling (MDS) algorithms have been verified to be robust for the mobile localization. However, they have not attained Cram?r-Rao lower bound (CRLB), even though the measurement noise is small. In this paper, a novel complex MDS method is proposed for time-of-arrival based mobile location. Unlike the existing MDS algorithms depending on the eigendecomposition of the scalar product matrix, the proposed method relies on the complex matrix and contains more information. We decompose a complex distance matrix using eigenvalue factorization and get a better performance. Computer simulations are included to contrast the estimator performance with several kinds of position methods.