Computed Tomography (CT)-based Finite Element (FE) models of femur have been extensively validated and recently applied to clinical studies showing encouraging results [1]. Nevertheless, there remains room for improvement in the accuracy of these models. Validation of FE models against in-vitro strain measurements and failure tests have identified systematic errors at sites where the cortical shell is thin (i.e. femoral neck) [2]. These errors may be explained by Partial Volume Effects (PVEs) in CT images [3, 4]. It has been suggested that CT deblurring algorithms can reduce PVEs in CT images, restoring both geometry and intensity data in thin bone structures [4]. The aim of this work was to investigate if a deblurring algorithm can improve CT-based FE modelling accuracy at the proximal femur.