Purpose: To evaluate the changes in the external-hexagon surface of the titanium (Ti) implant before and after mechanical cycling, when coupled with zirconia (Zr) abutments (A) manufactured by three computer-aided design/ computer-aided manufacturing (CAD/CAM) systems (Neodent Digital, Zirkonzahn, and AmannGirrbach) and the ZrTi abutment manufactured by Neodent. Materials and Methods: Four groups were formed (n = 6): titanium implant with Zr AmannGirrbach abutment (AZrAG), with Zr Zirkonzahn abutment (AZrZ), with Zr Neodent abutment (AZrN), and with Zr abutment with infrastructure in Ti Neodent (AZrTiN). Standardized abutments were made from three identical abutments milled in wax. Images of the surface of each side of the hexagons of the implant were obtained by scanning electron microscopy, before and after mechanical cycling, to evaluate the parameters: (1) scratches in the hexagon face; (2) hexagon superior shoulder kneading; (3) hexagon shoulder wear; (4) alterations on the hexagon base; and (5) scratches on the hexagon top. The abutments were coupled with the implants, and Cr-Co crowns were cemented. The implant/abutment/crown assemblies were submitted to mechanical cycling (400 N, 8.0 Hz) for 1 million cycles. The observed changes were classified as follows: absence (0), mild (1), moderate (2), and severe (3). The results were analyzed using the Mann-Whitney, Kruskal-Wallis, and Dunn tests (P .05). Results: For parameter 1, a significant difference (P = .008) was observed between AZrZ and AZrAG, with more scratches in AZrZ; and between AZrN and AZrTiN (P = .006), with more scratches in AZrN. For parameter 2, a significant difference (P .05) was observed between AZrZ and AZrAG and between AZrZ and AZrN, with greater kneading in AZrZ; among AZrN and AZrTiN, there was no significant difference (P = .103). For parameter 3, a significant difference (P .05) was observed between AZrZ and the other groups of Zr, with more wear in AZrZ; between AZrN and AZrTiN, there was no significant difference (P = .107). For parameter 4, a significant difference (P .05) was observed between AZrZ and AZrN, with more scratches in AZrZ; a significant difference (P = .002) was also observed between AZrN and AZrTiN, with more scratches in AZrN. For parameter 5, a significant difference (P .05) was observed between AZrZ and AZrAG and between AZrZ and AZrN, with the fewest scratches in AZrZ; a significant difference (P = .006) was also observed between AZrN and AZrTiN, with more alterations in AZrN. Considering all the alterations, the AZrZ group showed more surface alteration, 1.74 (0.99); followed by AZrN, 1.43 (0.92); AZrAG, 1.32 (0.96); and AZrTiN, 0.88 (0.94). Conclusion: Among the Neodent abutments, the AZrN group had shown more surface alterations. Among the Zr groups, AZrZ samples had shown the most altered surfaces, suggesting that alterations on the implant/Zr abutment hexagon surfaces are related to the abutment milled hexagon shape. Parole chiave: CAD/CAM, cyclic loading, esthetic abutment, implant abutment, implant-abutment interface, prosthodontics, titanium, zirconia