Bone regeneration is, like many other healing events, a complex, well-orchestrated process involving a myriad of different cell types and regulated by countless biochemical, physical and mechanical factors. But unlike other adult biological tissues, the majority of bone fractures can heal without the production of scar tissue, eventually recovering the original bone shape, size and strength. Despite bones remarkable healing capacity and the continuing research efforts, the impaired healing of complex orthopaedic cases is still not fully understood. This PhD work hypothesises that computational modelling can make a substantial contributionto the bone regeneration field by proposing and testing the underlying mechanisms of action as well as by designing and optimising experimentalstrategies in silico .In the first part of this work, an existing bioregulatory model of fracture healing is extended with an intracellular module of Dll4-Notch1 signalling in order to capture the ingrowth of new blood vessels through sprouting angiogenesis. The predictions of the new MOSAIC model are compared to experimental results and an extensive sensitivity analysis is performed on the newly introduced parameters. The potential of the MOSAIC model to investigate the influence of the molecular mechanisms on angiogenesis and consequently the bone formation process is illustrated.In the second part of this work, the MOSAIC model is further improved with a rigorous implementation of the influence of oxygen on the behaviour of skeletal cells. A comprehensive literature study is performed in order to ensure the correspondence of the oxygen ranges of the cell-specific oxygen-dependent processes with the state-of-the-art experimental knowledge. The oxygen model is corroborated with previously published experimental results. The robustness of the oxygen model with respect to the newly introduced oxygen thresholds is demonstrated by a sensitivity analysis. Some limitations and shortcomings of the oxygen model are identified together with suggestions for future work.In the last part of this work, the added value of the oxygen model is shown by applying it to three cases of impaired bone healing: the occurrence of nonunions in critical size defects, bone graft healing in a compromised environment and the impaired healing of bone fractures in NF1 patients. Not only is the oxygen model used to determine the underlying mechanisms of action, potential treatment strategies for the respective challenging orthopaedic conditions are also designed and optimised in silico .In conclusion, this PhD thesis demonstrates the potential of an integrative in vivo - in silico approach to advance our current understanding of bone regeneration as well as to design effective treatments of complex bone fractures.