Animal models of arthritis are widely used to de-convolute disease pathways and to identify novel drug targets and therapeutic approaches. However, the high attrition rates of drugs in Phase II/III rates means that a relatively small number of drugs reach the market, despite showing efficacy in pre-clinical models. There is also increasing awareness of the ethical issues surrounding the use of animal models of disease and it is timely, therefore, to review the relevance and translatability of animal models of arthritis. In this paper we review the most commonly used animal models in terms of their pathological similarities to human rheumatoid arthritis as well as their response to drug therapy. In general, the ability of animal models to predict efficacy of biologics in man has been good. However, the predictive power of animal models for small molecules has been variable, probably because of differences in the levels of target knockdown achievable in vivo.3