We examined 420 patients with vestibular diseases of different origin; 273 with peripheral vestibular disease and 147 with both peripheral and central vestibular disease. Recurrent vestibulopathy like Menière's disease, or benign paroxysmal positional vertigo, were not included. Patients were evaluated initially and 6 months after pharmacological and/or rehabilitation therapy. At the initial assessment, the head-shaking test was specific for the side of the lesion in both groups, even if spontaneous nystagmus was no longer present. Thus, head-shaking nystagmus is a physical sign that can be easily evoked and gives useful information about the presence of vestibulo-ocular reflex asymmetry. At the follow-up at 6 months, many changes in the head-shaking nystagmus were noted: in some cases it appeared, in some others it changed direction and more often it disappeared. There is actually no acceptable explanation for the disappearance of the head-shaking nystagmus, despite some evidence that vestibular compensation could play a role. It is definitely proved that sensitivity of the head-shaking test is really poor, especially in the course of time and, therefore, it should not be used alone in the follow-up of patients with vestibular disease.