Osteoarthritis (OA) is a major cause of suffering for millions of people. Investigating the disease directly on humans may be challenging. The aim of the present study is to investigate the advantages and limitations of the animal models currently used in OA research. The animal models are divided into induced and spontaneous. Induced models are further subdivided into surgical and chemical models, according to the procedure used to induce OA. Surgical induction of OA is the most commonly used procedure, which alters the exerted strain on the joint and/or alter load bearing leading to instability of the joint and induction of OA. Chemical models are generated by intra-articular injection of modifying factors or by systemically administering noxious agents, such as quinolones. Spontaneous models include naturally occurring and genetic models. Naturally occurring OA is described in certain species, while genetic models are developed by gene manipulation. Overall, there is no single animal model that is ideal for studying degenerative OA. However, in the present review, an attempt is made to clarify the most appropriate use of each model.