Abstract: The parameters can be estimated via two-dimensional (2D) multiple signal classification (MUSIC) algorithm, in which the two-dimensional spectral peak searching, however, requires much higher computational complexity. Aiming at this, an algorithm of root-MUSIC-based joint TOA and DOA estimation in IR-UWB system was proposed. After modeling the received signals in the frequency-domain, the algorithm estimated the TOA parameters first, and then the DOA parameters via the difference of the TOA was obtained, thereby the joint TOA and DOA estimation was gained. This algorithm does not need spectral peak searching and can obtain the closed-form solutions of the TOAs and DOAs. Moreover, it can get the parameters paired. In addition, the error variances of the estimated parameters were derived. The simulation results show that the proposed algorithm has much better performance than matrix pencil algorithm, propagator method and estimation of signal parameters via rotational invariance techniques (ESPRIT) algorithm, and has almost the same performance as 2D-MUSIC algorithm, while the complexity of the proposed algorithm has far lower than that of 2D-MUSIC algorithm.