Bisphosphonates are the most common pharmaceutical used to reduce fracture risk in postmenopausal women suffering from osteoporosis, but long-term bisphosphonate treatment has been associated with a rare adverse effect, atypical femoral fractures (AFFs). There is currently no suitable animal model to study the underlying pathophysiological mechanisms of AFFs, which remain unknown. Therefore, the objective of this project was to investigate a rabbit ulnar loading model for AFFs in cortical bone during long-term bisphosphonate treatment. The fatigue life and mechanical degradation of rabbit ulnae were measured during ex vivo loading of whole forelimbs in order to determine allowable limits for in vivo fatigue loading regimens. Post-fatigue mechanical testing of whole ulnae was determined to be impractical for use in the model due to the number of animals required and low sensitivity to fatigue damage. Therefore, a novel method was developed for measuring the plane strain fracture toughness of small animal bones at multiple locations along the diaphysis using arc-shaped tension specimens. This test method was able to produce highly repeatable measurements in replicates (coefficient of variation < 15%) and detect differences in fracture toughness due to small differences in tissue mineral density. Finally, an in vivo rabbit ulnar loading model was implemented to study AFFs in cortical bone during long-term bisphosphonate treatment. Several low body mass rabbits suffered displaced fractures during ulnar loading and were thus removed from the study. Treatment groups included a sham (control) and ovariectomized rabbits which were administered either vehicle (saline), a clinical dose, or a high dose of bisphosphonate (alendronate) over one year. Differences in tissue mineral density and fracture toughness between treatment groups were not statistically significant. Preliminary results from histology indicated that periosteal bone formation was stimulated in all treatment groups immediately following in vivo ulnar loading and returned to negligible basal levels within three months, as expected. Interestingly, intracortical remodeling (labelled osteons) was increased immediately following in vivo ulnar loading in the vehicle treatment group and not in the bisphosphonate treatment groups, but further analysis is needed.