In this paper, we consider a channel quantization problem that selects the best codeword representing a group of consecutive subcarriers (cluster) in multiuser MIMO-OFDM systems. The clustering is required to reduce the amount of feedback information in practical OFDM systems. Based on $M$-PSK based codebooks, we propose a new codeword search algorithm for the clustered MIMO-OFDM with complexity of O(2^{N_T}) compared to the complexity of O(M^{N_T}) for conventional exhaustive search, where denotes the number of transmit antennas. Through simulations, the proposed clustered codeword selection scheme for the multiuser MIMO-OFDM shows up to 54% performance improvement in throughput compared with the conventional MIMO-OFDM system that performs the channel quantization only on predetermined pilot subcarriers.