Because of the high cost and long time frame of clinical testing, animal models play a crucial role in the identification and selection of agents for the treatment of osteoporosis. The use of animal models early in a program focuses on the establishment of efficacy, while animal models used later in a program to examine bone safety. More specifically, animal models are used to gain information on the skeletal mechanism of action, to examine multiple skeletal sites (axial and appendicular), and to examine the effects of higher doses than will be used in humans. Animal models also predict the usefulness of surrogate markers in clinical trials, such as formation and resorption markers, as well as bone density. The hazard of using surrogate markers for fracture prevention is highlighted by high dose fluoride administration, which can increase bone density (considered a strong predictor of fracture protection) while not protecting against fractures. Estrogen-deficient models are most commonly used to mimic the postmenopausal bone loss in women; these models are characterized by increased bone turnover and a negative bone balance. The timing of the administration of the new therapy in animal models can help determine whether the agent will be more effective in the prevention of osteoporosis or in the treatment of established osteoporosis. New methods for the measurement of bone mass or volume are less invasive, require shorter acquisition time, and have enhanced resolution, resulting in increased knowledge concerning architectural changes and specific sites of bone deposition. Finally, the measurement of biomechanical strength of bones from animal models can be used to predict protective effects on fracture rates in clinical trials. When used in combination with other methods, animal models can greatly increase our understanding of the pathophysiology of osteoporosis and can expedite the development of new therapies.