This report describes a patient with iron deficiency in bone marrow examination and iron-responsive restless legs syndrome (RLS), in whom serum ferritin levels were well above the conventional cutoff for considering iron deficiency. The predictive value of serum ferritin for iron deficiency in RLS depends on the cutoff employed, on the pre-test likelihood of iron deficiency and on coexisting inflammatory conditions. Bone marrow examination is helpful when ferritin levels are equivocal.