Purpose: The aim of this study was to assess the effects of simulated axial loadings on the pull-out force required to disengage a two-piece conical connection abutment from an implant. Materials and Methods: Ten conical connection abutments (Ankylos Regular/X Abutment, Dentsply-Friadent) and 10 implants (Ankylos C/X Implant) were used. The implant-abutment assemblies were divided randomly into two groups: control group (C) and experimental group (E). For group E (n = 5), a cyclic load of 18 to 180 N at a frequency of 10 Hz to 10⁶ cycles was applied centrally and along the long axis of the implant, whereas for group C (n = 5), each sample was put on a workbench without cyclic loading at the same time during the testing period. Before mechanical loading, the initial torque values and the total lengths of tested samples of groups C and E were recorded. After cyclic loading, the postloading reverse torque value, the total length, and the pull-out force of tested samples of groups C and E were recorded. The difference between the initial torque value and the postloading reverse torque value was defined as the total torque loss. The difference between the initial total length and the postloading total length was defined as the axial displacement. The data were analyzed by independent t test. Results: The mean pull-out force of the experimental group was 77.60 N (SD = 6.16 N), which was significantly larger than that of the control group (mean = 55.28 N; SD = 9.41 N) (P .05). The mean total torque loss and the mean axial displacement of the experimental group were both significantly higher than those of the control group (P .05). Conclusion: Simulated axial loading increased the pull-out force of loaded abutments in comparison with unloaded abutments. Under simulated axial loading, the pull-out force of abutments tended to increase as the axial displacement of abutments and the total torque loss of abutment screws were both increased. Keywords: axial loading, conical connection, implant abutment, pull-out force