Animal models have proved to be of considerable importance in elucidating mechanisms underlying joint damage in osteoarthritis (OA) and providing proof of concept in the development of pharmacologic and biologic agents that may modify structural damage in the OA joint. The utility of animal models in predicting the response to an intervention with a drug or biologic agent in humans, however, can be established only after evidence is obtained of a positive effect of the agent in humans. To date, no agent has been shown unequivocally to have such an effect, although diacerhein and glucosamine have recently been reported to lower the rate of loss of articular cartilage in patients with hip OA and knee OA, respectively, based on measurements of the rate of joint space narrowing in plain radiographs. Furthermore, the predominant manifestation of OA – and the feature that leads people with radiographic changes of the disease to decide to seek medical attention and contributes to the enormous medicoeconomic and socioeconomic burden imposed by the disease – is joint pain. Notably, none of the animal models of OA is a good indicator of the analgesic effects of pharmacologic agents. Indeed, it should not be assumed a priori that reduction in the rate of progression of joint damage in OA will be associated with a reduction in joint pain.