An important mode of fracture treatment is the internal fixation of complicated bone fractures with plates and screws or with bone screws. A particularly common fracture is the distal radius fracture. For unstable fractures, internal stabilization with a volar locking plate has become a standard method. However, insufficient anchoring of the screw in the bone can lead to loosening. Simulations of screw-bone constructs allow research questions to be addressed at relatively low cost and without the need for valuable tissue samples such as are necessary in experiments. In particular, simulations with finite element (FE) models have gained popularity. Two commonly used types of FE models are micro finite element models (µFE) and homogenized finite element models (hFE). µFE models are accurate and relatively easy to create, but require more computational power. hFE models, on the other hand, offer a computationally efficient modeling approach. However, trabecular bone microarchitecture is not considered (mm element size) and material properties are estimated based on homogenized material properties. This study investigates the relation of peri-implant volume-average strain energy density (SED) between hFE and µFE of single-screw bone constructs. The comparison should be made with different hFE modelling strategies under different load cases. Nine µCT scans of distal sections of the radius served as the basis of this study. A cylindrical bone sample was taken virtually from the radius sections and a screw was virtually implanted. Two different load cases (axial pull-out and shear) were analyzed for each bone sample. After solving the FE models, custom scripts were used to evaluate the output. The output included: 1) stiffness, 2) strain energy densities averaged in a cylindrical volume around the screw. The output was evaluated for each of the nine bone samples, both load cases and both model types (hFE, µFE). In a clinical scenario (i.e., including both trabecular bone and cortex) of a single screw-bone construct at the distal radius, a significant correlation was found between the mean strain energy densities (SED) of the peri-implant volume and between that of the spring stiffness of hFE and µFE models in two load cases. hFE models overestimated stiffness and underestimated volume-average strain energy densities. The difference between hFE and µFE showed a dependence on bone morphometric parameters and was particularly high in samples with low bone volume fraction. A sub-study on the modeling strategy of the trabecular bone material in the hFE models showed that the local orthotropy of the trabecular bone improved the accuracy and precision of the prediction of the SED distribution only slightly. Overall, this study showed that hFE models can predict volume-averaged peri-implant SEDs of screw bone constructs in good agreement with µFE results, but this agreement can degrade dramatically in bone samples with low bone volume quality.