Finite element analysis (FEA) based on CT datasets of the spine or hip or on high-resolution peripheral CT datasets of the distal forearm or tibia is now widely used in research and clinical trials to estimate bone strength. Its clinical potential has recently been endorsed by the International Society of Clinical Densitometry Zysset et al. (J Clin Densitom 18(3):359–92, 2015). In vitro validation studies demonstrated the superiority of FEA over DXA for the prediction of ultimate load. In vivo studies confirmed the superiority in the spine, but data were less conclusive in the hip and forearm. Here, in addition to low bone strength the risk of falling is a major determinant of fracture risk. The next level of FEA dissemination, the integration into clinical practice, still faces a number of challenges such as access to dedicated FE software and its integration into the clinical workflow. Also compared to DXA, current FEA techniques have not shown a consistent superiority for hip fracture prediction, while hip CT is associated with a higher radiation exposure than hip DXA. For many clinicians, FEA and the direct measurement of strength instead of BMD are a novel perspective. However, the increasing use of abdominal and pelvic CT scans initially obtained for other clinical diagnosis, for the secondary use to assess osteoporosis and fracture risk (opportunistic screening), may accelerate the use of FEA. In this contribution, the basic technical aspects and limitation of FEA are discussed and the clinically relevant outcome measures are presented. Further advanced topics will broaden the understanding of the various aspects of FEA. Afterward a summary of in vivo studies using FEA for fracture prediction is given, which also includes a discussion of the clinical value of FEA for bone strength measurements.