A study was conducted to investigate the hypothesis that mechanical loading of implants and the consequent stress and strain fields influence bone modeling and remodeling at the bone-implant interface. Two implants ad modum Brånemark were placed in each of 20 canine tibiae, allowed to heal for I year, and then subjected to a controlled loading protocol. Implants in the left limb were loaded in axial tension with a triangular waveform (300 N maximum, 10 N minimum, 330 N/s) for 500 cycles per day for 5 consecutive days; implants in the right limb served as unloaded controls. Twelve weeks after loading, polished undecalcified thick sections were examined with light and scanning electron microscopy to provide bone modeling and remodeling data, including the surface area of periosteal and endocortical modeling, the percentage of mineralized tissue in the bone threads, and the frequency of occurrence of preloading and postloading fluorochrome bone labels. Also, a three-dimensional finite element model was developed to investigate the strain state in the bone near loaded implants. The morphometric data were statistically analyzed in terms of individual load-control pairs and showed the following trends for loaded implants: (1) a net bone loss near the coronal portion of the implant, (2) a smaller percentage of mineralized tissue in the cortex, and (3) a decreased frequency of occurrence of postloading fluorochrome bone label in the cortex adjacent to the implant. The finite element model indicated regions of high strain on the periosteal surface adjacent to the loaded implants. The results support the premise that the bone loss observed around the neck of the loaded implants at 12 weeks postloading was a consequence of bone modeling and remodeling secondary to bone microdamage caused by the loading protocol. This scenario, as well as certain other features of the bone response at the interfaces, can be interpreted in light of existing bone modeling and remodeling theories that relate bone activities and mechanical loading. Schlagwörter: bone modeling, bone remodeling, interface, microdamage, overload