The explosive growth of wireless communication is creating the demand for highspeed,reliable,and spectrally efficient communication over the wireless medium.This can be achieved by adopting multiple-input multiple-output (MIMO) andorthogonal frequency-division multiplexing (OFDM) technologies.However,MIMO-OFDM systems have the weakness of the OFDM to channel time-variation.Thus,channel estimation is a major problem in multi-user MIMO-OFDM systems.Manymethods can be used for channel estimation but in this thesis we will propose a newmethod of Spline function interpolation called CR-spline function.This methodreduces the computational complexity significantly and avoids long processing delays.Simulations are carried out for both 2x2 and 2x3 MIMO-OFDM systems and forboth QAM and QPSK modulation methods.A comparison of SNR versus BER forboth Spline and LS method shows better results for Spline method.