In this paper, we investigate the Tangshan M7.8, Datong M5.9, and Zhangbei M6.2 earthquake groups, each of which includes three to four earthquakes of the same magnitude. The case study results of these earthquake groups in north China show that stations near earthquake epicenters can record a significant rate of earth resistivity anomalies, which are characterized by a wide spatial distribution and long duration. We applied the wavelet transform method to analyze the geo-resistivity data observed prior to the occurrence of the earthquake groups to study the wavelet power geo-resistivity spectrum and its relative variation. The results suggest that the wavelet power spectrum can well reflect changes in the geo-resistivity, and is an effective method for extracting short-term geo-resistivity anomalies. In the mid-term stage prior to earthquakes, the wavelet power spectra of geo-resistivity are abnormal. The relative wavelet power spectra show obvious increases prior to moderate earthquakes or an earthquake group, which can well identify short-term and impending changes in geo-resistivity.