Work which has been undertaken in the field of predictive heat transfer applied to IC engine cooling galleries is reviewed. The focus for the review is the need to use accurate and representative heat transfer models within current and future engine design tools. Theoretical modelling has not yet been realised owing to the tack of knowledge associated with the generation and growth of bubbles from surface nucleation sites. The transferability of data and associated empirical models, between one rig and another, is sensitive to the method used to replicate the metal-coolant interface and duct geometry. The work of Chen and others is reviewed with the aim of highlighting the applicability and accuracy of current empirical models.