To assess whether plate bending at a hole significantly changes the biomechanical properties of a locked screw. Coronal plane bends of 5-, 15-, or 45-degree angles were placed in 3.5-mm locking compression plates with the apex at a locking hole. An additional 45-degree angle test group was created in which a threaded screw head insert was placed before bending. Ten plates were tested in each group and compared with nonbent controls in a stepwise cyclic loading protocol. Statistically significant differences in protocol survival were shown between the control group and the 15-degree angle (P = 0.006) and 45-degree angle (P = 0.0007) groups. An apparent decrease in protocol survival in the 5-degree angle group did not reach statistical significance (P = 0.17). The average number of cycles survived was significantly different between the control group and the 15-degree angle (P = 0.027) and 45-degree angle (P = 0.0002) groups. The mean cycles to failure for the 5-degree angle group was 16% lower than for controls but did not reach statistical significance (P = 0.37). The test group bent to an angle of 45 degrees after placement of a threaded screw head insert showed no difference in protocol survival or in mean number of cycles survived compared with the regular 45-degree angle group. Bending of a 3.5-mm locking compression plate by more than 5 degrees at a locking hole results in a statistically significant decrease in survival of the corresponding locked screw. This effect cannot be prevented by the placement of a threaded screw head insert before bending.