Purpose: To compare abutment screw loosening in 24-degree angulation-correcting and straight implants subjected to nonaxial cyclic loading.Materials and methods: Seven external connection 24-degree angulation-correcting implants (AI) and seven external connection straight implants (SI) were embedded in acrylic resin within a brass housing. A hemispherical titanium fatigue abutment was secured to each implant using a titanium abutment screw tightened to 32 Ncm. Each implant-abutment complex was positioned within a tooth wear machine and subjected to 1,000,000 cycles of 50-Ncm nonaxial loading to simulate 1 year of function. The abutment screw removal torque values were measured before and after cyclic loading, and the differences were statistically analyzed using two-way ANOVA and post hoc pairwise Dunn tests. Scanning electron microscopy and finite element analyses were performed to assess the wear of the abutment screws.Results: The mean torque loss for the AI group was 21.44% (P < .001) compared to 24.56% (P < .001) for the SI group. There was a statistically significant difference between the AI and SI groups (P = .006).Conclusion: Both groups exhibited significant abutment screw loosening. Within the limitations of this study, 24-degree angulation-correcting implants resisted screw loosening significantly more than straight implants.