Restless legs syndrome (RLS) is a chronic sleep motor disorder that affects up to 10% of the general population. Except for periodic leg movements (PLM), which can be found in the great majority of RLS patients, no objective hematochimic or neurophysiological markers are available to prove the diagnosis, which is based on clinical standard criteria. Nowadays, the aetiopathogenesis of the syndrome is unknown. In a consistent sample of patients affected by the idiopathic form, the disease is inherited as an autosomal dominant trait related to an unidentified locus, while each symptomatic form is probably linked to a specific cause. Although of possible different origins, both the primary and secondary forms may share the same pathogenetic mechanism, which, even if unclear, could be characterised by a neurological dysfunction of the dopaminergic system. Several issues, including strong efficacy of dopamine-agonist treatments, support this theory, which is currently considered the main pathogenetic hypothesis. Most of the past studies tried to clarify the RLS mechanism using the neurophysiological, biochemical and neuroimaging techniques applied to the field of human research. Now the time has come to accept the challenge in creating an animal model of RLS, which may emerge as a decisive step in understanding RLS pathogenesis, and to develop and test new therapies. Even though there have been a few significant efforts, a valid animal model of RLS still does not exist. In past pioneering studies, the authors attempted to induce restless motor behaviour in animals by different strategies: antidopaminergic pharmacological interventions, spinal or cerebral lesions of specific regions involved in the motor control and in dopamine regulation, and selective deletion of genes coding for dopamine receptors. Rodents (mice and rats) were always chosen by the authors as the animals for their experiments. The current tendency in achieving an RLS model is generally represented by simulation of a symptomatic condition of RLS or by a direct interference of the dopaminergic system. In this regard, the pharmacological method had the intention to reproduce the neuroleptic-induced acathisia, the spinal lesional model was based on the hypothesis of myelopathic- related PLM, and the hypothalamic lesion tested the motor consequence of A11 dopaminergic neurons. Preliminary studies are underway to replicate the pregnancy-related form of RLS by using a hormonal intervention, and the iron-deficiency secondary form by using specific iron-free diets. Today, modern technologies are available to easily replicate in animals most of the symptomatic RLS conditions. In addition, more than a few well validated animal models of different diseases known to be related to RLS or PLM, for instance, Parkinson's disease, rheumatoid arthritis and renal failure, could also be exploited in addressing this topic. The real obstacle in achieving an RLS model is the absence of a certain diagnostic marker to recognise if the animal that underwent the different experimental procedures has developed the RLS condition or not. Concerning this issue, possible specific endpoints are represented by the increase in locomotor activity, which are ascertainable by different techniques, such as openfield or run-wheel activity, or by sleep fragmentation, in which the circadian shift can be verified by applying polysomnography on the animal. PLM are probably the only specific and reliable markers available to recognise and quantify experimentally induced RLS. Despite a few authors who reported the presence of limb-phasic, pseudoperiodic activity during sleep in old or in lesioned rats, the existence of spontaneous or provoked PLM in animals is still debated. Eventually, the PLM features in an animal could be markedly different compared to human ones. To recognise and characterise PLM in animals, three more essential steps are required: a method to record directly, as in humans, the activity of the tibialis anterior (TA) muscles, a consistent amount of normative control data on the TA activity in healthy animals, and reliable analysis to distinguish the generic phasic muscular activity to a possible unambiguous PLM pattern. This review includes a summary and a critical discussion of the previous tentative RLS models, proposals for other possible animal models, and firstly the preliminary normative data on TA activity during sleep in normal rodents.