This paper describes a battery model for hybrid electric vehicle applications. The model is able to predict battery performance and thermal behavior in response to inputs such as drive and test profiles, including 40-season equivalent key life tests (KLTs). In order to balance the dynamic driving simulation responsiveness and accuracy with computing efficiency over lifetime duration simulations, the present model is built on a semi-empirical basis with sufficiently detailed descriptions of multiple chemical-physical behaviors in effect during cycling and storage. A set of computing processes are running concurrently to address real-time power capability, response, and ageing, which are represented by cumulative key stress factors. This paper focuses on model validation against actual KILT data, and it meanwhile describes the importance of KILT for hybrid and electric battery implementation. The present model is also designed to offer functionality for incorporation into vehicle-level simulations in a Matlab/Simulink environment. Copyright (C) 2009 John Wiley & Sons, Ltd.