Abstract: The real-time monitoring of power cable insulation status is of great significance to ensure its safe and stable operation. To address the limitations of existing cable insulation online monitoring technology, in this paper we propose an online monitoring method for XLPE cable insulation based on low-frequency signal induction. Firstly, the basic principle of this method is introduced by an analysis on the low-frequency induction signal parameters and and optimization of the sampling methods. Then, an experimental platform is built to carry out online insulation monitoring for the artificially accelerated thermal aging cable and actual out-service cable. The results indicate that a signal, characterized by the voltage amplitude of 5 V and the signal frequency of 10 Hz, can be injected through the open triangle side of the potential transformer at the neutral point of the distribution network. This injection induces the zero-sequence voltage and current in the cable line, enabling accurate online measurement of cable insulation resistance, leakage current and dielectric loss factor. For the artificially accelerated thermal aging cable, the maximum error between the dielectric loss factors obtained by the low-frequency signal induction method and the offline polarization-depolarization current method is only 12.6%, and the results obtained by the two methods exhibit identical time-varying patterns. For the actual out-service cable, the test results of the low-frequency signal induction method are consistent with the actual operation status of the cable. The test results demonstrate that the low-frequency signal induction method can achieve accurate online measurement of cable electrical parameters, thereby satisfying the practical requirements of engineering application.