Optimal design and operation of food refrigeration processes requires knowledge of thermophysical properties including density, specific heat capacity (and hence enthalpy), thermal conductivity and initial freezing point.  Measurement of such properties is time-consuming and expensive given the wide range of foods and variations in composition and structure. Simple methods to predict properties from composition data are desirable. Density and heat capacity are volume and mass weighted-averages of the food components respectively.  Thermal conductivity is a volume-based property but also depends on structure. Further, an important consideration is the effect of air voids contained with the food or packaging system and ice for frozen foods. The recommended method uses a parallel model for components other than ice and air; the Levy model for ice and an Effective Medium Theory model for air voids. Comparisons against experimental data for a range of products and worked examples show the accuracy and utility of the methods.