The vestibulo-ocular reflex (VOR) induces a compensatory movement in the eye when the head is rotated, to maintain stable vision when we move. It originates in the peripheral vestibular system, which detects head movements. It is particularly effective for rapid head movements, as tested in the Head Impulse Test (HIT). In acute unilateral vestibular deficit (AUVD), the VOR deficit is compensated for by a substitution saccade, more commonly known as catch up saccade, that contribute to refocus the gaze and maintain vision during head rotations. Recent technological advances have made it possible to make high-quality recordings during HIT (video Head Impulse Test, vHIT), leading to the identification of substitution saccades of variable latency. Our team has shown that saccades of shorter latency lead to better visual function (Hermann et al., 2017) and that the cerebellum is involved in the development of these saccades (Hermann et al., 2023), suggesting a learning effect rather than the de novo appearance of particular saccades. The main hypothesis of this study is that the mechanisms underlying short-latency substitution saccades, which seems to guarantee good functional recovery, depend on learning occurring from the first days after an acute unilateral vestibular deficit. We also hypothesise that early physiotherapeutic rehabilitation of the VOR under Head Impulse Test conditions would promote this learning process and the development of early catch-up saccades. One of the causes of AVD is the resection of cochleovestibular schwannomas. This procedure involves a neurotomy, i.e. complete vestibular deafferentation, which is precisely known due to the scheduled nature of the surgery. The exact moment of onset of vestibular damage is therefore known, unlike other vestibular pathologies. Hospitalisation is necessary in the immediate aftermath of surgery, with the presence of physiotherapists on the wards. In addition, there is no spontaneous recovery of the vestibular deficit. These patients therefore represent the ideal acute unilateral vestibular deficit model for testing our hypothesis. Two studies using vHIT in the aftermath of vestibular schwannoma resection surgery (Pogson et al. 2022; Mantokoudis et al. 2014) also allow us to confirm the safety and feasibility of our protocol in this patient population.