OBJECTIVE: To investigate the relationship between the different defect length of vessels and the options of vascular repair, and to compare the different options of repair because of the longitudinal biomechanical effect. METHODS: A clinical analysis was undertaken to evaluate the major arterial and venous injuries in human extremities repaired by end-to-end anastomoses or venous autograft(177 cases, 185 vessels). Compared the defect length of the same kind of vessels repaired by different options (Student-t test). Evaluated the defect length to repair arterial injuries between by end-to-end anastomoses and by vein graft by means of 95% confidence interval. RESULTS: There was significant difference between the defect length of brachial artery repaired by end-to-end anastomosis and femoral artery and popliteal artery repaired by autogenous vein graft (P lt; 0.01). The upper limit of confidence interval in the defect length of brachial artery, femoral artery and popliteal artery was 3.17 cm, 2.81 cm and 2.44 cm respectively by end-to-end anastomosis by means of 95% confidence interval. The lower limit of confidence interval in the defect length of brachial artery, femoral artery and popliteal artery was 2.82 cm, 2.41 cm and 2.17 cm respectively by vein graft by means of 95% confidence interval. The defect length of brachial artery, femoral artery and popliteal artery repaired by vein graft was linear correlation with the length of graft. CONCLUSION: Because of the longitudinal biomechanical difference of arteries and veins in human extremities, different options of repair are necessary to different arterial injuries.