Quantitative analyses of hemimandibular kinematics during mastication have not yet received much attention during the past two decades as the technology for measuring the kinematics have greatly improved. Chew cycle kinematics and inferred tooth occlusion have been applied to hemimandibles as if they were a fused single mandibular system, or as if they were a half mandible in 2-D lateral view, despite the well-established understanding that hemimandibles are rendered semi-independent by their semi-mobile symphysis. Treating the hemimandibles as relatively independent and mobile can have major implications for interpreting functional morphology of the fossils of early mammals. The research of this thesis measures hemimandibular kinematics during mastication in six Didelphis virginiana individuals eating two food types, almonds and cheese (Chapter 1). A traditional coordinate axis is applied to the system to measure the pitch, yaw, and roll of each hemimandible, as well as a novel helical axis analysis. Here, I find that for description and quantification of hemimandibular kinematics the four chew cycle phases previously defined and deployed in chewing studies are more appropriately divided into six chew cycle phases: fast close 1 (FC1), fast close 2 (FC2), slow close 1 (SC1), slow close 2 (SC2), slow open (SO), and fast open (FO). No one dimension of rotation predominates or is “more important” during slow close. Roll and yaw, especially, vary in relative magnitude depending on the side of an individual and between individuals. This variation has important implications for the degree of stereotypy of the hemimandibular system. The novel helical axis method provides valuable understanding the whole mandibular system. Importantly, the rotational data show that the axis of rotation was found to be outside the working side hemimandible for the majority of the chew cycle. Hemimandibular mastication is fundamentally different from mastication in a fused mandible and this thesis represents a critical first step in precisely measuring and defining kinematics of the hemimandibular system.In the second half of my thesis, I studied the dentoalveolar neurofeedback system during mastication by transecting the left inferior alveolar nerve, while leaving the right one intact. Overall, the kinematics measured during post-transection mastication were found to be a mix of individual learned responses to nerve transection and a stereotyped change to the entire system. A couple of patterns emerged: (1) Both hemimandibles show changes in kinematics after the transection, even when the healthy right side is the working side. (2) Transected individuals preferentially evert either their right or left hemimandible for the entire cycle. (3) Roll is the rotational dimension with the most change, and mediolateral translation of the M1 talonid basin is the translational dimension with the most change. These results suggest that the phases of the chew cycle where tooth-tooth or tooth-food-tooth contact are likely to occur (end of FC2, SC1, SC2, and start of SO) are controlled bilaterally through dentoalveolar neurofeedback. Transecting the inferior alveolar nerve unilaterally does not entirely inhibit mastication, but limits the ability of the animal to respond to changes in food physical properties in all dimensions, primarily roll. This is relevant for interpreting the early mammalian evolution as the earliest mammaliaforms had at least partially mobile symphyses. The hemimandibular kinematics in Didelphis virginiana are particularly important for interpreting the functional evolution of the therian mammals, in which changes in the tribosphenic molar are coupled with changes to the jaw shape, such as the angular region. The opossums of didelphid marsupials have been used as proxies to infer function of the Mesozoic therian mammals. For an early therian mammal that has a tribosphenic molar with complex occlusal surfaces, it is a pre-requisite that it also has a neurofeedback system, for coordinated and effective function of the mandible and teeth.