The purpose of this study was to develop a finite element model (FEM) of a human orbit, who experienced a pure orbital blowout fracture, to study the effect of the geometrical mismatch-induced stresses on the orbital floor/graft interface and how to improve the graft design when restoring the orbital floor. A FEM of the orbit and the globe of 1 patient who experienced pure orbital blowout fracture and treated with autogenous bone graft was generated based on computed tomographic scans. Simulations were performed with a computer using a commercially available finite element software NISA (EMRC, Troy, MI). The FEM was then used to study the effects of changing the geometry, position, material properties, and method of fixation of the autogenous bone graft on its predictions. The factors that had the biggest impact on the predicted principal strain magnitudes were absence of cancellous bone (up to 60%) and bony support of the graft (up to 50%). Applying rigid fixation reduced stresses by 30% posteriorly and by almost 100% anteriorly. Alterations to the geometry of the bone graft, such as an increase in its thickness, increased principal strain magnitudes (up to 42%). Applying rigid fixation reduced principal stresses significantly. The role of rigid fixation becomes more prominent when there is no bony support posteriorly and/or medially. This study also highlights the importance of preserving cancellous bone, when harvesting and preparing the autogenous bone graft to reconstruct the orbital floor. The possibility that absence of cancellous bone and the resulting stresses may be a source of graft resorption and/or failure cannot be excluded.