To accurately predict internal tissue loads for early diagnostics of diabetic foot ulcerations, a novel data-driven computational analysis was conducted. A dedicated dual fluoroscopic system was combined with a pressure mat to simultaneously characterize foot motions and soft tissue’s material properties during gait. Finite element (FE) models of the heel pad of a diabetic patient were constructed with 3D trajectories of the calcaneus applied as boundary conditions to simulate gait events. The tensile and compressive stresses occurring in the plantar tissue were computed. Predictions of the layered tissue FE model with anatomically-accurate heel pad structures (i.e., fat and skin) were compared with those of the traditional lumped tissue (i.e., homogeneous) models. The influence of different material properties (patient-specific versus generic) on internal tissue stresses was also investigated. The results showed the peak tensile stresses in the layered tissue model were predominantly found in the skin and distributed towards the circumferential regions of the heel, while peak compressive stresses in the fat tissue-bone interface were up to 51.4% lower than those seen in the lumped models. Performing FE analyses at four different phases of walking revealed that ignorance of layered tissue structures resulted in an unphysiological increase of peak-to-peak value of stress fluctuation in the fat and skin tissue components. Thus, to produce more clinical-relevant predictions, foot FE models are suggested to include layered tissue structures of the plantar tissue for an improved estimation of internal stresses in the diabetic foot in gait.