Animal models have been used for a number of decades to study arthritis and have contributed greatly to unravelling mechanisms of pathogenesis and validating new targets for treatment. All animal models have sets of limitations and over the years there has been natural refinement of existing models as well as creation of new ones. The success of genetic modification in mice has fuelled an exponential increase in the use of murine models for arthritis research and has significantly increased our understanding of disease processes. This review focuses on those rodent models of RA and OA that have current utility and are widely used by the research community. We highlight the subtle but important differences in existing models by positioning them on a pathogenesis map whereby model selection is determined by the specific aspect of disease to be studied. We discuss the evolving challenges in in vivo arthritis studies and our perceived gaps for future new model development. The document includes technical and cost implications of performing the described models, and the ethical considerations of such approaches.