In this paper, we present a unique word (UW)-based channel estimation approach for multiple-input multiple-output (MIMO) systems under doubly dispersive channels, which is applied to orthogonal time frequency space (OTFS) with space time coding (STC). The OTFS modulation has been recently proposed as a robust technique under time varying channels due to its property of spreading the data symbols over time and frequency. Yet another relevant aspect is the employment of multiple antennas at the transmitter and receiver. Therefore, we consider an STC MIMO system with cyclic delay diversity at the transmitter and maximum ratio combining at the receiver, where we develop a UW-based channel estimation scheme for multiple transmit antennas. We show a recently proposed frame optimization scheme for SISO is directly applicable to MIMO. In addition, we evaluate numerically the frame error rate (FER) of OTFS and OFDM with 2x2 and 4x4 MIMO, where the time varying channel is estimated using the UW-based approach. The FER results reveal that OTFS becomes more advantageous than OFDM for MIMO-STC systems with higher order modulation and code rate.