In this paper, the waveform design for Rician target detection with multiple-inputmultiple-output radar is addressed. The relative entropy between the distributions of the observations under two hypotheses is employed as the design metric. Due to the difficulty of tackling the non-convex waveform design problem, a novel algorithm based on minorisation-maximisation is developed. Since a simple quadratic function is established to minorise the objective function, the presented algorithm has lower computational complexity than its counterparts. Moreover, it can be extended to design waveforms under many practical constraints, including the constant-envelope constraint, the similarity constraint, and the constant-envelope and similarity constraints. Numerical results are provided to show the effectiveness of the proposed algorithm for detecting Rician targets. Moreover, the designed waveforms can be used to enhance the detection performance of a Swerling 0 target in the presence of angle mismatch.