Abstract-By exploiting the micro-motion features of fast rotating targets, wideband radar has been successfully applied to high resolution imaging. However, due to the traditional fixed pulse repetition interval (PRI), the target image may suffer from aliasing in some practical situations. In this paper, under the compressed sensing (CS) radar framework, an efficient wideband imaging scheme with random PRI signal is introduced for aliasing reduction. Considering that direct application of the CS theory will result in large-scale dictionaries and high computational complexity, we firstly generate a low resolution image by applying modified generalized Radon transform on range-slow time domain and then scale down the dictionary column by reserving the atoms corresponding to those strong scattering areas. Simulation results show that this scheme can achieve aliasing-free images with acceptable computational cost.