Accurate heat release analysis of cylinder pressure data is a powerful tool used in the development of diesel engines. However, significant errors in the calculated heat release values can occur due to shortcomings in both the experimental measurements and in the heat release model and this can produce misleading results. This paper shows the effect of such common errors on the calculated gross heat release data obtained when analysing simulated and experimental direct injection diesel engine pressure diagrams using a traditional single-zone First Law heat release model. The work reveals that the greatest uncertainty in most cases will be caused by assuming the wrong rate of heat transfer between the cylinder charge and combustion chamber walls. To overcome this limitation, an alternative heat release model is proposed and shown to give very good results over a wide range of operating conditions. This heat release model, which uses a variable polytropic index to cater for the heat transfer, is best suited for diesel engine performance and development applications where consistent diesel engine heat release data are required.