We performed torsional resistance tests to examine the Protaper Retreatment D1, D2 and D3 and the NRT-GPR 3N and 4N, which are Ni-Ti files used to remove gutta-percha in root canal retreatment. On cyclic fatigue failure tests, the D1 was the quickest to break, and the 4N took the longest to break. Torsional bending tests showed the fracture angle to be largest for the 4N, which was significantly different from that of all the other files (p < 0.01). There were no significant differences among the D1, D2 and D3. Torsional torque tests showed that the superelastic D1, that has the largest diameter, had the largest fracture torque. Files with larger diameters had higher fracture torques. On bending torque test, the D1 differed significantly from all the other files (p < 0.01). Significant differences were also seen between different versions of the same file type (p < 0.01). All of the fracture surfaces of files used in torsional resistance tests in the present experiment showed SEM findings that appear to be instantaneous breakage.The results of this experiment indicated that torsional resistance test values varied according to differences in file properties and shape. It may be necessary to consider these results when using files in clinical practice.