If the active interference enters the main lobe of a radar antenna, the detection performance of the radar system will be severely degraded. Many traditional techniques for suppressing this type of interference are not sufficiently effective. Through the analysis of the characteristics of linear frequency modulated echoes and jamming in the fractional Fourier transform (FrFT) domain, this paper proposes a novel algorithm for moving ground target detection, which can eliminate multipath signals effectively and suppress the main-lobe jamming significantly. First, by choosing the right pulse number taken into the accumulation, the effect of the active interference and moving target echoes can be suppressed simultaneously in the fractional domain, and then the FrFT of multipath signals can be obtained. Second, the FrFT of moving target echoes can be extracted by differential cancellation, and the second coherent integration is employed to detect the moving targets and construct the FrFT-Doppler plane. By constant false-alarm rate detection, the peaks can be extracted from the FrFT-Doppler plane, and then the target parameters can be estimated properly. It is not necessary to estimate the direction of arrival of each interference or limit the number of interferences in the main lobe. The processing results of simulated and actual datasets verify that for the weak or low-velocity targets, the detection performance of the proposed algorithm is superior to that of conventional match filter processing.