The invention discloses a distance-Doppler imaging method based on multi-task Bayes compression sensing and mainly solves problems of high complexity and low resolution existing in the prior art. The method comprises steps that 1), a pulse modulation sequence is emitted by a radar to a target scene, and an echo signal is further received by the radar; 2), a target scene distance scope and a Doppler scope are divided at equal intervals, discrete sampling for the echo signal is carried out, and an echo matrix is constructed; 3), an auxiliary variable is cited, and an optimization solution function is constructed; and 4), the optimization function is solved, and the distance-Doppler imaging result of the scene is acquired. According to the method, scene distance-Doppler dimensions are reconstructed, not only are algorithm complexity reduction and resolution improvement realized, but also complex parameter setting is not required, and the method can be applied to fields of baffle field monitoring, missile guidance and space detection.