Class I foods with the lowest maximum thermal conductivity ratio are the simplest foods to predict thermal properties for since the uncertainty involved is relatively low, as indicated by the small region bounded by the Series and Parallel models (Fig. 2a). In this case, most thermal conductivity models commonly found in the food engineering and refrigeration literature will provide predictions of sufficient accuracy. However, food Classes II, III and IV provide greater challenges to thermal conductivity prediction since the thermal conductivity of ice is an order of magnitude higher than