Hydrogen-fuelled internal combustion engines are being investigated as an alternative for current drive trains because they have a high efficiency, near-zero noxious and zero tailpipe greenhouse gas emissions. A thermodynamic model of the engine cycle would enable a cheap and fast optimization of engine settings for operation on hydrogen, facilitating the development of these engines. The accuracy of the heat transfer submodel within the thermodynamic model is important to simulate accurately the emissions of oxides of nitrogen which are influenced by the maximum gas temperature. These emissions can occur in hydrogen internal combustion engines at high loads and they are an important constraint for power and efficiency optimization. The most common heat transfer models in engine research are those from Annand and Woschni. These models are developed for fossil fuels, which have different combustion properties. Therefore, they need to be evaluated for hydrogen. We have measured the heat flux and the wall temperature in an engine that can run on hydrogen and methane. This paper describes an evaluation of the models of Annand and Woschni, using those heat flux measurements and assesses if the models capture the effect of changing combustion and fuel properties. The models fail on all the tests, so they need to be improved to accurately model the heat transfer generated by hydrogen combustion. (C) 2010 Professor T. Nejat Veziroglu. Published by Elsevier Ltd. All rights reserved.