Abstract This correspondence discusses over-the-air (OTA) testing for multiple-input multiple-output (MIMO) capable terminals, with an emphasis on test zone size characterization for multi-probe anechoic chamber (MPAC) OTA configurations. For an MPAC setup, it is important to understand the test zone size that can be supported by its design. Further, it is desirable that the test zone size should be determined in terms of measured throughput deviation. However, such works have not been reported in the literature yet, due to challenges in practical MPAC setups. In this correspondence, we propose to simulate the MPAC OTA configuration and MS design in a channel emulator and perform MIMO throughput measurements in a conductive setup. With the proposed scheme, we can investigate how large a test zone can be supported for the synthetic MPAC configuration in terms of throughput deviation. This scheme is attractive, since various MS antenna designs and MPAC configurations can be flexibly selected and test zone size in terms of measured MIMO throughput deviation can be achieved. Measurement results show that with a high antenna correlation at the base station (BS) side, measured throughput results would be low, irrelevant to MPAC OTA designs and antenna designs at the MS side. Therefore, the antenna correlation at BS side should be set uncorrelated to ensure that true MS performance can be measured. Further, spatial correlation accuracy at the MS side becomes critical for throughput accuracy only when correlation values are in the high region (e.g., ρ > 0.5).