The delay of fracture healing may expose elderly persons to the higher risk of osteoporosis or, in the worst case, put them into a bedridden state. Angiogenesis, a requisite for fracture healing, is regulated by angiogenic factors such as vascular endothelial growth factor (VEGF), and its expression is mediated by hypoxia inducible factor (HIF). This study was undertaken to test the hypothesis that stabilizing HIF expression reduces the unloading-induced delay of bone repair through the enhancement of angiogenesis. Angio- and osteogenesis within a rat cortical bone defect at early healing stages were examined by k-edge subtraction CT using synchrotron light and a newly developed zirconia-based vascular contrast-casting agent (Zr-CA). Under pentobarbital anesthesia (40 mg/kg ip), rats (♀, 12 wk) receiving a drill-hole surgery on a tibial diaphysis were divided into three groups undergoing no treatment (C), hindlimb unloading by tail suspension (HU), and HU with desfferioximine administration (HU-DFO). Desfferioximine, an agent to stabilize HIF, was injected close to the defect site (200 μM, 20 μL) from postoperative day 1 on alternate days. At postoperative day 5 (DAY5: C, n=10; HU, n=13; HU-DFO, n=11) or 10 (DAY10: C, n=10; HU, n=11; HU-DFO, n=11), each rat was anesthetized again (pentobarbital, 40 mg/kg ip) and perfused with Zr-CA from the abdominal aorta. The rats were then euthanized with KCl administration (3 M, 0.5 ml iv) and immersed in ice-cold water for solidifying Zr-CA. The defect site was scanned at SPring-8 (Harima, Japan) with 17.9 and 18.1-keV X-ray, below and above the zirconium k-edge, respectively, and the two scan data sets were reconstructed with 2.74-μm voxel resolution. Taking advantage of contrast enhancement of vascular image at 18.1 keV, vascular and bone images were obtained separately by subtraction between 17.9 and 18.1-keV images after 3D-registration based on mutual information. Volume fractions of bone (B.Vf) and vasculature (V.Vf) were calculated, and values (means±S.E.M.) were compared between groups at each healing stage by the Kruskal-Wallis test followed by Dunn’s multiple comparison test. At DAY5, B.Vf [%] was higher in HU-DFO (1.50±0.38) than in HU (0.28±0.08) (p<0.05) although both values were lower than B.Vf in C (3.60±0.78) (p<0.05). V.Vf [%] tended to be lower in HU (4.48±1.31) than in C (7.37±2.00) but did not differ between HU-DFO (8.30±1.17) and C. At DAY10, bone occupied a larger space than vasculature. B.Vf was lower in HU (26.0±2.14) than in C (43.9±1.93) (p<0.05) but similar between HU-DFO (39.5±1.21) and C. V.Vf did not differ between HU-DFO (6.22±0.78) and HU (5.01±0.74) and both values were lower than V.Vf in C (8.10±0.92) (p<0.05). These results suggest that poor angiogenesis at a very early stage of healing is involved in the delay of bone defect filling under hindlimb unloading and that the enhancement of angiogenesis through the HIF-VEGF pathway reduces this delay of bone defect healing.