This paper presents a preamble-based time domain SNR estimator for an adaptive modulation of STBC assisted MIMO-OFDM system employing \(M\)-PSK. The modified Constant Amplitude Zero Autocorrelation (CAZAC) preamble structure is exploited in the preamble-based time domain SNR estimator, which exhibits estimated SNR difference of about 0.09 dB compares to the actual SNR for SUI-5 channel. It is observed that the difference between the estimated SNR using the CAZAC-MIMO SNR estimator and the actual SNR is smaller compared to that of the estimated SNR using the benchmarking SNR estimator and its performance is further improved as the diversity order of the system increases. The proposed adaptive modulation of STBC assisted MIMO-OFDM system achieves BER and channel throughput improvement, benefited from the employment of the proposed SNR estimator. At BER\(=10^{-4}\), the proposed adaptive modulation system outperforms the corresponding non-adaptive system employing 8-PSK by about 2 dB. Meanwhile, the performance comparison between the proposed adaptive modulation system and the corresponding adaptive modulation of single-input single-output (SISO)-OFDM system demonstrated significant BER and channel throughput improvement, benefiting from spatial diversity of \(2 \times 2\) MIMO-OFDM system.