AbstractObjectivesThe purpose of this study was to evaluate the axial displacement of the abutments during clinical procedures by the tightening torque and cyclic loading.Materials and methodsTwo different implant‐abutment connection systems were used (external butt joint connection [EXT]; internal tapered conical connection [INT]). The master casts with two implant replicas, angulated 10° from each other, were fabricated for each implant connection system. Four types of impression copings were assembled and tightened with the corresponding implants (hex transfer impression coping, non‐hex transfer impression coping, hex pick‐up impression coping, non‐hex pick‐up impression coping). Resin splinted abutments and final prosthesis were assembled. The axial displacement was measured from the length of each assembly, which was evaluated repeatedly, after 30 Ncm torque tightening. After 250 N cyclic loading of final prosthesis for 1,000,000 cycles, additional axial displacement was recorded. The mean axial displacement was statistically analyzed (repeated measured ANOVA).ResultsThere was more axial displacement in the INT group than that of the EXT group in impression copings, resin splinted abutments, and final prosthesis. Less axial displacement was found at 1‐piece non‐hex transfer type impression coping than other type of impression copings in the INT group. There was more axial displacement at the final prosthesis than resin splinted abutments in the INT and the EXT groups. After 250 N cyclic loading of final prosthesis, the INT group showed more axial displacement than that of the EXT group.ConclusionInternal tapered conical connection demonstrated a varying amount of axial displacement with tightening torque and cyclic loading.