The mutual coupling and gainphase errors of the transmitting and receiving arrays would significantly degrade the performance of the estimation algorithms for highresolution direction of arrival (DOA) and direction of departure (DOD) in bistatic multiple input multiple output radar. To solve this problem, a joint estimation algorithm for twodimensional (2D) angle of the targets and the mutual coupling and gainphase error matrices of the transmitting and receiving arrays is proposed, which exploits the subspace theory and reduced dimension idea by applying several wellcalibrated instrumental sensors in both transmitting and receiving arrays. Firstly, DOA and DOD of the targets can be estimated accurately by the algorithm without the knowledge of the mutual coupling and gain phase error matrices. Secondly, an accurate estimation can also be achieved for the mutual coupling and gain phase error matrices based on the accurate estimation of the 2D angle for the targets. Furthermore, the selfcalibration is realized for the array errors. The proposed algorithm only needs one dimensional spectral peak search, and does not need high dimensional nonlinear optimized search, so it has a little calculation. The efficiency and accuracy of the proposed algorithm are verified by the computer simulation results.