The self-sensing high ductility cementitious composite that integrates high ductility and sensing performance has a unique advantage in structural health monitoring. The study of the electrical conductivity is significant for analyzing the sensing performance. In this paper, the high ductility self-sensing cementitious composite was developed by adding carbon fibers (CFs) into high ductility cementitious matrix. The percolation threshold and the effects of curing age, moisture content and temperature on the conductivity of the material were analyzed. The conductive behavior was also analyzed by equivalent circuit model and electrical impedance spectroscopy (EIS). The results revealed that the percolation threshold is 0.7 wt%. When the CF content is lower than the percolation threshold, the conductivity decreases with the increase of curing age, but increases with the increase of moisture content and temperature. The material conductive behavior is caused by the joint action of ionic conduction and electronic conduction of conductive fillers. When the CF content is higher than the percolation threshold, the effects of curing age and temperature are weakened, and the moisture effect has disappeared. The electronic conduction dominates the conductive behavior. It was also concluded that the equivalent circuit model can be applied to describe the conductive behavior.