Non-unions remain a clinical problem and are characterised by the failure to heal after a defined period of time. Current preclinical non-union models apply a wide variety of techniques to diminish intrinsic healing potential deviating from the clinical situation. The aim of this study was to develop and characterise a non-union model in rats using internal plate fixation without the need for additional healing insults, whereby bone healing can be longitudinally assessed using microCT. It was hypothesized that healing/non-unions can be accurately predicted at early time points by microCT. Female, skeletally mature Fischer F344 rats received a 2mm or 1mm femoral osteotomy, stabilized with either a 2mm thick plate or a 1.25mm thick plate. Healing was monitored by microCT over 14weeks and histological analysis at euthanasia. The mechanical environment was characterised using finite element (FE) modelling and biomechanical testing. The majority of animals receiving the 2mm thick plate displayed poor healing responses in both the 2mm and 1mm defect size groups. Bone and cartilage formation were markedly improved using the 1.25mm thick plate. MicroCT could accurately predict bone forming capacity at early time points (3-4weeks). The 2mm thick plating system confers poor healing responses in female Fischer F344 rats, comparable to atrophic non-unions. By reducing plate thickness to increase interfragmentary strain within the defect site healing is improved, leading to borderline healing situations or increased abundance of cartilage tissue present in the defect site with ultimate failure to bridge the defect (hypertrophic non-union). Furthermore, microCT can reliably identify delayed/non-healing animals within 4weeks, thereby allowing their selective targeting for the testing of novel, clinically relevant treatment strategies in different clinical situations aimed at restoring impaired bone healing.