The usual method to prefabricate a bone flap is to harvest a nonvascularized bone graft and to implant the artery and vein bundle between segments of bone graft. The basic problem of this method is sacrificing an artery for prefabrication. Another method for creating flap donor sites without using an artery is venous flap prefabrication. There are a few articles describing bone flap prefabrication, and these include implantation of both artery and vein as a vascular bundle. Also, there is no experimental study in the literature using a vein or an arterialized vein pedicle for bone flap prefabrication. As an experimental model for bone flap prefabrication, the rabbit ear vascular model was chosen. For the experiments 3 groups were formed. Each group contained 5 rabbits. In the first experimental group a vein was implanted between the halves of bone graft. In the second experimental group an arterialized vein was implanted between the halves of bone graft. To compare the viability of the bone graft of the 2 prefabrication groups, a bone graft was implanted into the subcutaneous pocket of the posterior auricular area in the third group. The authors examined 5 rabbits in each group by microangiography at the end of 6 weeks except for group 3. On microangiographic analysis, groups 1 and 2 showed patency of the vascular pedicle. There was no difference between these 2 groups from the point of view of vascular patency and bone appearance. Bone scintigraphy was performed for 5 rabbits in each group. On bone scintigraphic scans, the bone component of the flaps was visualized in groups 1 and 2, but not in group 3. A quantitative analysis of images was performed by drawing symmetric spherical regions of interest (ROIs) over both the implanted area and cranial bone. The uptake ratios were computed by dividing the mean counts in the implanted ROI by mean counts in the cranial bone ROI. The mean value was 0.86 +/- 0.02 in group 1 and 0.86 +/- 0.04 in group 2. A statistically significant uptake difference was not seen between venous and arterialized venous groups (P < 0.01). Histologic examination was performed all rabbits in each group, and demonstrated that the bony component was viable, showing osteocytes containing lacunae, osteoblasts along bony trabeculae, and vascular channels in groups 1 and 2. In group 3, the bony architecture of the graft was still apparent, but all bone within it was dead. There were no significant microangiographic, histologic, and scintigraphic differences between the 2 experimental methods.