In order to tackle the issue of poor consistency of internal individual parameters in retired electric vehicle batteries, an SGA-FCM amended fuzzy C-means clustering method is proposed. Firstly, select sorting parameters based on the second-step Thevenin equivalent RC circuit model of lithium battery cells to construct a retired battery sorting model and choose the terminal voltage, remaining capacity along with equivalent total resistance of battery cells as the sorting arguments. Secondly, a simple genetic algorithm is introduced to preprocess the initial clustering centers to achieve global optimal sorting based on the sensitivity of traditional FCM to outliers. The final simulation results of algorithm running show that the sorting efficiency and accuracy can be improved in some degree by the SGA-FCM, and the static and dynamic properties of the new battery modules after sorting also meet the national standards requirements, which means that the consistency of batteries can be effectively guaranteed.