To improve the crash safety of battery electric vehicles, thermal runaway of lithium-ion cells should be avoided during crash scenarios. Consequently, a predictive approach should be developed. In this study, we propose a constitutive model for predicting mechanically induced short circuits that could lead to thermal runaway. To fulfil the compromise between high accuracy and computational performance, an accurate deformation model is required. In this study, prismatic EV32 cells were investigated. Due to the complexity of the layered inner structure of Li-ion battery cells and for computational performance reasons, a simplified continuum model is necessary for full vehicle crash simulations. The layered structure of the inner cell, consisting of a periodic stack of anodes, cathodes and separators, suggests the use of a transversely isotropic material model. Furthermore, the large proportion of active material covering each electrode — especially the graphite covering the anode — necessitates a compressible formulation. This is achieved by incorporating some weighted hydrostatic stress as proposed by Benzerga and Besson [2] into the compressible model developed by Deshpande and Fleck [1]. The model uses an associated flow rule and two strain-hardening functions that depend on both the deviatoric and hydrostatic stress components, as well as a Johnson–Cook strain rate dependency [3]. A short-circuit criterion depending on local deformation and strain rate has been defined. The model has been implemented as a user subroutine in the LS-DYNA FE code. The cutting plane algorithm from Ortiz and Simo (4) is used to integrate the model. Calibration and validation of the proposed model are based on experimental tests of prismatic battery cells, including compression and indentation tests with various punch velocities and different indenter geometries. The model accurately captures the results of the experimental tests, particularly the earlier occurrence of a short circuit at increasing velocities.