Fats and oils are found in many food products; however, their macroscopic properties are difficult to predict, especially when blending different fats or oils together. With difficulties in sourcing specific fats or oils, whether due to availability or pricing, food companies may be required to find alternative sources for these ingredients, with possible differences in ingredient performance. Mathematical and computational modeling of these ingredients can provide a quick way to predict their properties, avoiding costly trials or manufacturing problems, while, most importantly, keeping the consumers happy. This review covers a range of mathematical models for triacylglycerides (TAGs) and fats, namely, models for the prediction of melting point, solid fat content, and crystallization temperature and composition. There are a number of models that have been designed for both TAGs and fats and which have been shown to agree very well with empirical measurements, using both kinetic and thermodynamic approaches, with models for TAGs being used to, in turn, predict fat properties. The last section describes computational models to simulate the behavior of TAGs using molecular dynamics (MD). Simulation of TAGs using MD, however, is still at an early stage, although the most recent papers on this topic are bringing this area up to speed.