Although multi-user MIMO multiplexing is known as a powerful means to increase the sum throughput, MIMO diversity still remains as an important means to improve the bit error rate (BER) performance in a poor channel condition. Recently, we proposed selective MIMO diversity with subcarrier-wise user equipment (UE) antenna identification/selection. Selective MIMO diversity jointly uses maximal-ratio transmit diversity frequency-domain equalization (MRTD-FDE) and minimum mean square error combining diversity FDE (MMSECD-FDE) both at base station (BS) for OFDM downlink and SC uplink, respectively. In this paper, we compare selective MIMO diversity with two other typical MIMO diversity schemes, i.e., the space-time block-coded (STBC) diversity and the MMSE filtering combined with singular value decomposition (MMSE-SVD) based diversity. Pilot-aided channel estimation (PACE) is used for generating the transmit/receive diversity equalization weights at BS and the transmit/receive diversity weights at UE. The uncoded average BER performance of selective MIMO diversity in a frequency-selective Rayleigh fading MIMO channel is evaluated by computer simulation and is compared with those of STBC-TD and MMSE-SVD diversity.