Both contaminated natural soil and simulated contaminated kaolinite have been used for electrokinetic remediation (EKR) of toxic metal(s). The objectives of this study are to make a comparison on the EKR performance between lead‐contaminated natural soil (NatSoil) and simulated lead‐contaminated kaolinite (KlnSoil), and to explore the intrinsic mechanism for the discrepancy between these two cases. Results show that the EKR process on NatSoil failed to achieve effective Pb enrichment in any soil section (in total four) with either KNO3 or EDTA‐2Na as the catholyte. However, enrichment ratios (C/C0) of 2.11 and 4.45 are achieved on KlnSoil with the above two electrolytes, respectively. The speciation of Pb in different soil samples, 55.81% of Pb in reducible fraction in NatSoil and 85.06% of Pb in exchangeable and soluble fraction in KlnSoil, are attributed to the significant differences of remediation. The mineral components, cation exchange capacity, acid/base buffering capacity are more complex than that of KlnSoil, hindering the desorption and migration of Pb2+. Increasing the voltage from 2 to 4 V cm−1, pre‐saturating the soil sample with EDTA‐2Na instead of deionized water and prolonging the treatment time from 240 to 480 h successfully enhance the enrichment ratio of Pb at the third soil section to 2.44, 1.70, and 1.69, respectively. The energy consumptions of the remediation processes are also evaluated. Overall, pre‐saturating the soil samples with EDTA‐2Na is the optimal strategy for NatSoil considering the balance between the remediation efficiency and energy consumption.