Purpose: The vertical location of the implant-abutment connection influences the subsequent reaction of the peri-implant bone. It is not known, however, whether any additional influence is exerted by different microgap configurations. Therefore, the radiographic bone reactions of two different implant systems were monitored for 6 months. Materials and Methods: In eight mongrel dogs, two implants with an internal Morse-taper connection (INT group) were placed on one side of the mandible; the contralateral side received two implants with an external-hex connection (EXT group). On each side, one implant was aligned at the bone level (equicrestal) and the second implant was placed 1.5 mm subcrestal. Healing abutments were placed 3 months after submerged healing, and the implants were maintained for another 3 months without prosthetic loading. At implant placement and after 1, 2, 3, 4, 5, and 6 months, standardized radiographs were obtained, and peri-implant bone levels were measured with regard to microgap location and evaluated statistically. Results: All implants osseointegrated clinically and radiographically. The overall mean bone loss was 0.68 ± 0.59 mm in the equicrestal INT group, 1.32 ± 0.49 mm in the equicrestal EXT group, 0.76 ± 0.49 mm in the subcrestal INT group, and 1.88 ± 0.81 mm in the subcrestal EXT group. The differences between the INT and EXT groups were statistically significant (paired t tests). The first significant differences between the internal and external groups were seen at month 1 in the subcrestal groups and at 3 months in the equicrestal groups. Bone loss was most pronounced in the subcrestal EXT group. Conclusions: Within the limits of this study, different microgap configurations can cause different amounts of bone loss, even before prosthetic loading. Subcrestal placement of a butt-joint microgap design may lead to more pronounced radiographic bone loss. Schlagwörter: bone morphology, crestal implant placement, implant-abutment microgap, radiographic study, subcrestal placement