The underlying basis of this research is that food oral breakdown heavily depends on the mechanical properties of the material being masticated. So far, researchers have been measuring mechanical properties to establish sensory panels which are useful towards designing food. Although this methodology has proven industrially popular, it displays weakness in addressing complex product development and optimisation tasks. Specifically, mastication involves varying parameters across consumers, such as speed, teeth geometry, bite force, jaw motion, as well as friction. Therefore, this study considers more advanced techniques to help predict how the food will perform during oral processing. A Finite Element (FE) model was developed, where the first bite on a starch based food was simulated, based on digital pet teeth geometry. The food material model was constructed from uniaxial compression and tension data as well as fracture toughness values through the essential work of fracture (EWF) and cutting tests. The main focus is on the development of suitable material constitutive laws including damage, in order to predict deformations and crack patterns in the food item, as experienced during chewing. Emphasis is also given to determining the true material fracture toughness by assessing the validity of the EWF and cutting methods on highly dissipative materials, which leads to the development of a new, more accurate and convenient test method. The composition effect on the mechanical response was also studied by comparing four starch based recipes. The FE jaw force versus displacement results match the experimental data obtained by a physical replicate of the bite model, lending weight to the approach as a powerful tool in understanding food breakdown and in establishing relations between the mechanical properties and sensory attributes. The study also reflects industry needs for time and cost efficient techniques towards product development and optimisation.