By exploring sparsity in the target space, compressive sensing (CS) based multi-input multi-output (MIMO) radar systems achieve either the same localization performance as traditional methods but with significantly fewer measurements, or significantly improved performance with the same number of measurements. The work presented here investigates the performance gain of CS-MIMO radars, stemming from optimal power allocation among the transmit antennas, or optimal waveform design. In both cases, the optimization criterion is the minimization of the coherence between the target returns from different search cells, or equivalently, the coherence of the columns of the sensing matrix.