Anorexia nervosa (AN) is a severe psychiatric disorder with high rates of morbidity, comorbidity and mortality, which in a subset of patients (21%) takes on a chronic course. Since an evidence based treatment for AN is scarce, it is crucial to investigate new treatment options, preferably focused on influencing the underlying neurobiological mechanisms of AN. The objective of the present paper was to review the evidence for possible neurobiological correlates of AN, and to hypothesize about potential targets for Deep brain stimulation (DBS) as a treatment for chronic, therapy-refractory AN. One avenue for exploring new treatment options based on the neurobiological correlates of AN, is the search for symptomatologic and neurobiologic parallels between AN and other compulsivity- or reward-related disorders. As in other compulsive disorders, the fronto-striatal circuitry, in particular the insula, the ventral striatum (VS) and the prefrontal, orbitofrontal, temporal, parietal and anterior cingulate cortices, are likely to be implicated in the neuropathogenesis of AN. In this paper we will review the few available cases in which DBS has been performed in patients with AN (either as primary diagnosis or as comorbid condition). Given the overlap in symptomatology and neurocircuitry between reward-related disorders such as obsessive compulsive disorder (OCD) and AN, and the established efficacy of accumbal DBS in OCD, we hypothesize that DBS of the nucleus accumbens (NAc) and other areas associated with reward, e.g. the anterior cingulated cortex (ACC), might be an effective treatment for patients with chronic, treatment refractory AN, providing not only weight restoration, but also significant and sustained improvement in AN core symptoms and associated comorbidities and complications. Possible targets for DBS in AN are the ACC, the ventral anterior limb of the capsula interna (vALIC) and the VS. We suggest conducting larger efficacy studies that also explore the functional effects of DBS in AN.