Rheumatoid arthritis (RA) and Osteoarthritis (OA) are both common diseases of the joints. RA is distinguished by inflammation and synovitis leading to joint destruction, whereas OA is typified by degenerative and mostly noninflammatory disease. Although differing in their pathology, both forms can elicit chronic disabling pain in patients. Relief from this pain is an unmet need for many patients, and this has lead to a drive to understand the pain mechanisms occurring in each disease in order to develop new analgesics. This necessitates the use of pre-clinical models. Here we discuss the methods and limitations of animal pain assessment, rodent models of RA and OA, and how these models have been used to investigate the genesis of pain. Specifically, we focus on processes studied in both RA and OA models, and how the mediators involved in the development of pain may differ between these two arthritic states and during acute and chronic pain. From this we have learnt that the key mediators of RA pain include inflammatory cells, inflammatory cytokines, astrocytes, substance P, and CGRP. Endogenous analgesic mediators in RA include β-endorphin, μ-opioid receptor, and various anti-inflammatory cytokines. Less is known of the mediators of OA pain, but important factors include CGRP, TRPV1, NGF, VIP and the μ-opioid receptor. Interestingly, several factors have been found not to play a role in OA pain, including glial cells, neutrophils and TNF. It must be noted that the vast majority of candidate drugs from animal research never reach the human clinic, in part due to false positives from animal models. This may be due to flaws in the models themselves, the methods used to assess pain, the physical properties of the candidate drug, or an inherent difference between animal and human pain pathways. By developing more clinically relevant models, novel diseasespecific analgesics are being developed with the hope of improving the quality of life for sufferers of arthritis pain.