Abstract: A new lowcomplexity user selection algorithm was proposed for multipleinput multipleoutput (MIMO) system employing block diagonalization (BD) with othercell interference (OCI) suppression. The algorithm iteratively selects users through greedy search. In each user selection step, the user who contributes most to the cell capacity is added to the selected user set, and the number of activated users is adaptively adjusted to maximize the cell capacity. The simulation results show that the proposed algorithm can achieve about 99% of the cell capacity of the complete search at high SNRs, and at low SNRs the two come out almost the same.