In wireless communication, the mismatch between the transmit and receive oscillators and Doppler effect result in carrier frequency offset (CFO) which has a big effect on communication systems. Estimation of frequency offset is becoming an important, urgent, and practical issue needed to be cared for, especially in multiple input-multiple output (MIMO) systems. The fact shows that an inaccurate frequency offset estimate will reduce the accuracy of other system parameter estimates, which diminishes the expected spectral efficiency of the MIMO antenna system. Regarding to this problem, two algorithm candidates are presented to estimate carrier frequency offset exactly in a MIMO system in this paper. While the first one focus on matrix algorithm, the second one which is more complicated is a subspace-based approach called MUSIC (MUtilple SIgnal Classification) to overcome the limit of DFT resolution.