Automotive radars make use of angle information obtained from antenna arrays to distinguish objects that lie in the same range-doppler cell (relative to the ego vehicle). This paper proposes novel ways of using presently known minimum redundancy arrays (MRAs) in single-input multiple-output (SIMO) and multiple-input multiple-output (MIMO) automotive radars. Firstly, an MRA-based sparse MIMO array is proposed as a novel modification to the nested MIMO array. The proposed sparse MIMO array uses MRAs as the transmitting and receiving modules, unlike the nested MIMO array, which uses two-level nested arrays (TLNAs) at the transmitting and receiving blocks. Upper bounds for the sum co-array aperture and the overall attainable degrees of freedom (DOF) offered by the MIMO radar have been derived in terms of the number of sensors. Secondly, the suitability of large Low-Redundancy Linear Arrays (LRLAs) in SIMO automotive radars is also studied. A long-range automotive radar driving scenario was assumed for DOA estimation and simulations were carried out in MATLAB using the co-array MUltiple SIgnal Classification (co-array MUSIC) algorithm. Simulation results confirm that the proposed MRA-based MIMO array provides better angular resolutions than the nested MIMO array for the same number of sensors and that LRLAs can serve as a handy replacement for ULAs in SIMO radars owing to their acceptable performance. As MIMO and SIMO radars designed from currently known MRAs were sufficient to satisfy the angular resolution requirements of modern automotive radars, a need to synthesize new MRAs did not arise. © 2020 Electromagnetics Academy.