Purpose: The objective of this study was to analyze orthotropic bone formation and remodeling of three different dental implant surfaces with and without recombinant human bone morphogenetic protein 2 derived from Escherichia coli (ErhBMP-2) in a rabbit model. Materials and Methods: Resorbable blasting media (RBM); sandblasted, large-grit, acid-etched (SLA); and magnesium-incorporated oxidized (MgO) implant surfaces were coated with ErhBMP-2 (1.5 mg/mL). The implants were placed into the proximal tibia in six New Zealand White rabbits. Each rabbit received six different implants (three coated with ErhBMP-2 in one tibia and three uncoated implants in the other tibia), and the sites were closed, submerging the implants. The animals received alizarin (at 2 weeks), calcein (at 4 weeks), and tetracycline (at 6 weeks) fluorescent bone markers, and were euthanized at 8 weeks for histomorphometric analysis. Results: The amount of ErhBMP-2 coating was 9.6 ± 0.4 µg per MgO implant, 14.5 ± 0.6 µg per RBM implant, and 29.9 ± 3.8 µg per SLA implant. Clinical healing was uneventful. Mean bone-to-implant contact (± standard deviation) for the ErhBMP-2/RBM (35.4% ± 5.1%) and ErhBMP-2/MgO (33.4 % ± 13.2%) implants was significantly greater compared with RBM (23.6% ± 6.2%) and MgO (24.9% ± 2.7%) implants (P .05). Considering the mean bone-to-implant contact in cortical bone, ErhBMP-2/SLA implants (32.9% ± 7.8%) showed lower bone-to-implant contact in cortical bone than all other implant variations (range, 39.9% ± 18.1% to 51.3% ± 9.2%; P .05). There were no remarkable differences in new bone area, with minor differences between implants. Conclusions: Within the limits of study, it was found that the absorbed ErhBMP-2 dose varied with implant surface characteristics, influencing local bone formation and remodeling.