Purpose: The passive fit of a superstructure on implant abutments is essential to success. One source of error when using a tapered cone-screw internal connection may be the difference between the tightening torque level applied to the abutments by the laboratory technician compared to that applied by the treating clinician. The purpose of this study was to measure the axial displacement of tapered cone-screw abutments into implants and their replicas as a function of the tightening torque level. Materials and Methods: Twenty tapered cone-screw abutments were selected. Two groups were created: 10 abutments were secured into 10 implants, and 10 abutments were secured into 10 corresponding implant replicas. Each abutment was tightened in increasing increments of 5 Ncm, from 0 Ncm to 45 Ncm, with a torque controller. The length of each sample was measured repeatedly with an Electronic Digital Micrometer. The mean axial displacement for the implant group and the replica group was calculated. The data were analyzed by the Mann-Whitney and Spearman tests. Results: For both groups, there was always an axial displacement of the abutment upon each incremental application of torque. The mean axial displacement values varied between 7 and 12 µm for the implant group and between 6 and 21 µm for the replica group at each 5-Ncm increment. From 0 to 45 Ncm, the total mean axial displacement values were 89 µm for the implant group and 122 µm for the replica group. Conclusion: There was a continuous axial displacement of the abutments into implants and implant replicas when the applied torque was raised from 0 to 45 Ncm. Torque applied above the level recommended by the manufacturer increased the difference in displacement between the two groups. Int J Oral Maxillofac Implants 2009;24:251–256