Purpose: The purpose of this study was to compare the screw joint stability between the CADCAM custom-made implant abutment and the prefabricated implant abutment by measuring the reverse torque value after cyclic loading. Materials and methods: Twelve screw type implants (Implantium, Dentium Co., Seoul, Korea) were embedded in aluminum cylinder with acrylic resin. The implant specimens were equally divided into 3 groups, and connected to the prefabricated titanium abutments (Implantium, Dentium Co., Seoul, Korea), CADCAM custom-made titanium abutments (Myplant, Raphabio Co., Seoul, Korea) and CADCAM custom-made zirconia abutments (Zirconia Myplant, Raphabio Co., Seoul, Korea). The CAD-CAM milled titanium crown (Raphabio Co., Seoul, Korea) was cemented on each implant abutment by resin cement. Before cyclic loading, each abutment screw was tightened to 30 Ncm and the reverse torque value was measured about 30 minutes later. After the crown specimen was subjected to the sinusoidal cyclic loading (30 to 120 N, 500,000 cycles, 2 Hz), postloading reverse torque value was measured and the reverse torque loss ratio was calculated. Kruskal-Wallis test was used for statistical analysis of the reverse torque loss ratio. Results: The CADCAM custom-made titanium abutments presented higher values in reverse torque loss ratio without statistically significant differences than the prefabricated titanium abutments (P>.05). Reverse torque loss ratio of the custom-made zirconia abutments was significantly higher compared to that of the prefabricated titanium abutments (P=.014). Conclusion: Within the limitation of the present in-vitro study, it was concluded that there was no significant difference in screw joint stability between the CADCAM custom-made titanium abutments and the prefabricated titanium abutments. On the other hand, the CADCAM custom-made zirconia abutments showed lower screw joint stability than prefabricated titanium abutments. (J Korean Acad Prosthodont 2012;50:128-34)