Internal short circuit (ISCr) detection of a battery is critical for preventing thermal runaway and enhancing electrical vehicle safety. In this paper, the electrical characteristics of the ISCr of a large format lithium ion battery are analyzed using the equivalent circuit model (ECM). An ISCr detection method is developed based on battery consistency within the battery pack. The ISCr detection method employs the recursive least square (RLS) algorithm based on the mean-difference model (MDM), which is derived from the ECMs consisting of the mean and difference value of cells' voltage and resistance. The algorithm first estimates the basic parameters of the MDM. Then the algorithm calculates the characteristic parameters, such as the differential of the voltage and the fluctuation function of the internal resistance, derived from the basic parameters of the MDM. These characteristic parameters obviously vary once ISCr occurs, thereby helping the battery management system to realize ISCr detection. The effectiveness of the ISCr detection method is validated through a series of experiments. (C) 2015 Elsevier B.V. All rights reserved.