An experimental study was performed using the canine sciatic nerve in order to compare the efficacies of a vascularized trunk graft (VTG), a free trunk graft (FTG), a vascularized cable graft (VCG), a free cable graft (FCG) and a two stage procedure (TSP) for the repair of a peripheral nerve defect. Both a 4 cm, and a 5 cm nerve defect was repaired using each method, and also a 6 cm defect, using VCG and FCG. After 24 weeks, the intraneural blood flow (IBF) was measured in the 4 cm defect group, and the recovery of nerve regeneration and muscle reinnervation were evaluated by motor nerve conduction velocity, wet muscle weight, and histological examinations of the nerve and the muscles. VTG and VCG both demonstrated superior IBF. The IBF after TSP was less than after VTG and VCG but greater than after FTG and FCG. In the 4 cm defect group, VTG showed the most favorable recovery in both the axonal regeneration and muscle reinnervation and was followed by VCG. Although the recovery after FTG, FCG and TSP appeared to be worse than after VTG and VCG, there was no significant difference. In the 5 cm defect group, VTG showed the most favorable recovery followed by VCG and FCG. For axonal regeneration, FTG was significantly worse than VTG, VCG and FCG. TSP was worst within the 5 cm defect group for muscular reinnervation, although it showed somewhat better axonal regeneration than FTG. In the 6 cm defect group, no significant difference was found between VCG and FCG for the axonal regeneration. Clinically, TSP is used for repairing a short nerve defect just beyond the critical distance that cannot be overcome by a primary end-to-end suture. Nerve grafting, other than FTG, appeared to be the most reliable method of bridging a long nerve defect, and FCG might be the most practical method.