Galvanic vestibular stimulation

Galvanic vestibular stimulation (GVS) is a non-invasive technique that applies a small electrical current through electrodes placed on the mastoid processes behind the ears to stimulate the vestibular system of the inner ear. The current changes the firing of the vestibular nerve afferents, so a person who is sitting or standing still can feel head motion, sway or tilt that is not happening. In their 2004 review, Richard Fitzpatrick and Brian Day described GVS as "a simple, safe, and specific way to elicit vestibular reflexes".

GVS has been used for more than a century as a tool for studying balance, eye movements and self-motion perception, and more recently in the clinic for vestibular and neurological disorders. For virtual reality (VR), it is of interest because a head-mounted display can show self-motion that the inner ear does not feel, and this visual-vestibular conflict is a leading explanation for simulator sickness and cybersickness. Researchers have tested GVS signals synchronized with the displayed motion to reduce that conflict, to add a sense of acceleration without a motion platform, and to strengthen vection and presence. Samsung and the Mayo Clinic licensee vMocion both announced GVS systems for VR entertainment in 2016.

Reviewed 6 October 2026. Checked every cited paper (authors, title, venue, year, DOI via Crossref and Europe PMC full texts or abstracts), the SIGGRAPH 2005 paper and archive page, the Samsung newsroom, Engadget and Mayo Clinic/vMocion releases against the claims they support. About review dates.

How it works

Electrode montages

In the most common setup, called binaural bipolar (or bilateral) stimulation, the anode is placed on one mastoid process and the cathode on the other, so the current flows through the skull past both vestibular organs. Other montages place one electrode on a mastoid and the second on the forehead, the vertex or the forearm. Two-pole stimulation behind the ears produces a sense of head roll; a three-pole setup with an added electrode on the forehead produces pitch. In 2015 Kazuma Aoyama and colleagues reported that a four-electrode montage using the mastoids and the temples could also induce a virtual head yaw motion and yaw body sway, giving directional control about three axes. For flight simulation, Michael Cevette, Jan Stepanek and colleagues used a multisite electrode array and identified six electrode combinations that produced motion perception in the three rotational axes.

Effect on the vestibular nerve

Cathodal current increases the firing of vestibular afferent fibers and anodal current decreases it; switching the current off produces the opposite change in firing. In recordings from awake macaque monkeys with electrodes placed behind the ears in the same binaural bipolar configuration used in humans, Kwan, Forbes, Mitchell, Blouin and Cullen found in 2019 that transmastoid GVS activates semicircular canal and otolith afferents in parallel, rather than mainly the otoliths as some earlier work had proposed. Irregularly firing afferents responded more strongly than regular ones, and the tuning of the afferents to GVS differed from their tuning to natural head motion. Fitzpatrick and Day noted in 2004 that evoked responses could not be interpreted with certainty because it was not understood how the stimulus acts as an input to the vestibular system.

Perceptual and postural responses

With binaural bipolar stimulation, a standing person shows a sustained body sway toward the anode that stabilizes within one to two seconds and returns to upright after the current stops. Seated or supine subjects report small oscillating sensations of roll (about 5 to 15 degrees in one study), while stronger currents have produced reports of larger rotations. GVS also evokes torsional eye movements, which are used as a measure of vestibular activation. Combined with a moving visual scene, GVS can change perceived self-motion: in a 2006 study of visually induced vection, trapezoidal GVS bent the perceived path of forward or upward self-motion sideways, toward the cathode side after stimulus onset.

Waveforms and current levels

Studies use constant (direct) current steps, square or pulse waveforms, sinusoids and noise. Reported intensities range from sub-milliampere noise to several milliamperes, and some ocular studies went up to 5 or 7 mA, though the 2025 review by Marchand and colleagues notes that clear ocular responses can be obtained below 1 mA and questions the need for high intensities. Noisy GVS (nGVS) applies a zero-mean random current, often at a level the user cannot perceive. It is thought to work through stochastic resonance, in which added subthreshold noise improves the performance of the peripheral vestibular system. In a 2014 study by Iwasaki and colleagues, white-noise GVS between 0 and 1,000 µA improved standing balance in 76 percent of 21 healthy subjects and 91 percent of 11 patients with bilateral vestibular dysfunction.

History

The effects of passing current through the head were described before they were linked to the vestibular system. In 1800 Alessandro Volta applied his newly invented battery to his ears and reported spinning, imbalance and a boiling sound. Purkinje described balance disturbances from currents through the head in 1819. In the 1870s Hitzig recorded nystagmus in response to electrical stimulation in dogs and humans, and Breuer combined galvanic stimulation with removal of the labyrinth in animals to show that the responses came from the vestibular organs. Both men tried the stimulation on themselves and reported a sensation of falling toward the side of the cathode. The modern method keeps the same principle but places the electrodes on the mastoids rather than in the ears.

Fitzpatrick and Day wrote in 2004 that despite its long history it had "only recently found popularity as a research tool and is rarely used clinically". By 2019 Kwan and colleagues described GVS as rapidly gaining popularity for the assessment and treatment of clinical disorders including Parkinson's disease, stroke, cerebellar ataxia, vestibulopathy and concussion. Noisy GVS is now the most common form in clinical settings.

GVS was also proposed as a human interface. At SIGGRAPH 2005 Emerging Technologies, Taro Maeda, Hideyuki Ando and Tomohiro Amemiya of NTT Communication Science Laboratories, with Naohisa Nagaya, Maki Sugimoto and Masahiko Inami of the University of Electro-Communications, showed a GVS device called "Shaking the World". It induced a virtual sense of acceleration synchronized with optic flow or music and could steer a walking person sideways toward the anode. The authors proposed it as a way to add acceleration to video "without an expensive mechanical motion platform" and noted that the electrodes could be built into headphone pads. Maeda and Ando were also co-authors of the 2015 four-pole study and of "GVS RIDE", which paired a head-mounted display with four-pole GVS and was shown at SIGGRAPH 2017 Emerging Technologies.

In 2016 GVS appeared in announcements of consumer VR products. On 14 March 2016 Samsung announced that it would show Entrim 4D, an experimental headphone-style accessory from its C-Lab program, at the SXSW festival in Austin. On 30 March 2016 the Mayo Clinic and vMocion announced that vMocion's "3v Platform" would bring the clinic's patented GVS technology to VR and augmented reality.

Applications in VR and AR

Reducing simulator sickness

One approach is to "recouple" the vestibular signal with what the user sees. In a 2012 study by Cevette, Stepanek and colleagues, GVS electrode combinations were first calibrated in 21 subjects, and subjects were then randomly assigned to fly a cockpit flight simulator with or without GVS synchronized to the speed and direction of the visual field. Self-reported simulator sickness scores in the GVS group were 6.3, 20 and 8 percent for gastrointestinal, central and peripheral symptoms, against 17, 22.4 and 20 percent in the control group. A 2022 study with several of the same authors integrated GVS with a VR flight simulator. With GVS, the 20 participants showed lower electrogastrogram measures (the dominant power instability coefficient and the percentage of bradygastric waves) than in the control session, better flight performance, and better secondary task performance on the easy task. The authors described the results as initial evidence.

Head-mounted VR studies have taken similar approaches. Misha Sra, Abhinandan Jain and Pattie Maes of the MIT Media Lab presented a small wearable GVS device at CHI 2019. An Android app sent the stimulus over Bluetooth Low Energy to electrodes behind the ears, and the system supported driving, flying, teleporting and riding experiences. In a study with 20 participants, cybersickness was significantly lower with the device, and participants rated the GVS experiences as more immersive than a no-GVS baseline. Groth and colleagues calibrated GVS signals in real time to the camera motion of pre-recorded 360-degree videos filmed from a moving first-person viewpoint, and reported in 2022 that this significantly reduced discomfort for cybersickness-susceptible users.

Noisy GVS has also been tested. Weech, Wall and Barnett-Cowan gave participants 30 minutes of noisy GVS (plus or minus 1,750 µA) or sham stimulation while they played one of two VR games. For the more nauseating game, cybersickness on a verbal scale was lower during and directly after stimulation, but questionnaire measures did not differ, the effect disappeared within about 3 to 6 minutes of further VR exposure, and there was no effect for the milder game.

Motion sensation, vection and presence

GVS can add a felt acceleration to visual motion, which is the basis of the 2005 NTT device and of later entertainment systems. In a 2026 study in the journal Virtual Reality, Peter Wagner, Stephen Palmisano, Rahm Ranjan, Shinichi Iwasaki and Juno Kim had standing participants view simulated forward self-motion with a 0.4 Hz side-to-side head oscillation in a headset. GVS at 0.4 Hz that matched the visual oscillation produced the strongest vection (22.1 percent higher than the same oscillating display without GVS) and raised spatial presence by 15.6 percent against that condition. GVS did not significantly change cybersickness in that study, and it reduced user comfort.

Flight and spaceflight training

GVS can also be used to create a vestibular error on purpose. In a 2011 study at the NASA Ames Vertical Motion Simulator, pseudorandom GVS served as an analogue of the sensorimotor effects of weightlessness during simulated Space Shuttle landings. Among 11 subjects, unsuccessful (crash) landings rose from 2.3 percent without GVS to 9 percent with it. In a 2022 VR flight simulation study with 19 participants, deliberately mismatched multi-axis GVS induced approximate somatogravic and Coriolis illusions, which the authors proposed as a way for pilots to experience flight illusions safely on the ground. Subjective nausea was low overall but significantly higher in the GVS session.

Balance, cognition and body perception

VR is also used as a test environment for GVS itself. A 2026 study with 30 healthy adults found that noisy GVS improved standing balance when a VR scene created visual-vestibular conflict. In a 2025 VR study of 32 healthy adults, Bhardwaj and Sra reported that noisy GVS significantly improved spatial memory performance compared with the condition without it. A study using the rubber hand illusion suggested that GVS can modify bodily awareness.

Commercial efforts

Year	Company	System	Description
2016	Samsung (C-Lab)	Entrim 4D	Headphone-style accessory shown at SXSW with a Samsung Gear VR racing-car demo. Samsung said the team had tested it on more than 1,500 people and developed 30 movement patterns, and described it as still in development.
2016	vMocion with Mayo Clinic	3v Platform	Exclusive, global, perpetual license to the Mayo Clinic's GVS patents and algorithms for media and entertainment, based on research by the clinic's Aerospace Medicine and Vestibular Research Laboratory.

Samsung described GVS in its announcement as "a safe and simple technique that sends specific electric messages to a nerve in the ear", and said its team was working on a version with additional electrodes to create a sense of rotational motion. vMocion stated that the 3v Platform synchronizes the vestibular and visual fields "within one-tenth of one second" and "eliminates VR sickness in most people"; these are company claims, not results from a published trial.

Research summary

Year	Authors	Venue	Setting	Reported result
2005	Maeda, Ando, Amemiya et al.	SIGGRAPH Emerging Technologies	GVS interface device	Vection synchronized with optic flow or music; walking guidance toward the anode
2012	Cevette, Stepanek et al.	Aviation, Space, and Environmental Medicine	Cockpit flight simulator	Lower simulator sickness scores with synchronized GVS
2015	Aoyama, Iizuka, Ando, Maeda	Scientific Reports	Four-pole electrode montage	Virtual head yaw in addition to roll and pitch
2019	Sra, Jain, Maes	CHI 2019	Wearable device with VR content, 20 participants	Lower cybersickness, higher rated immersion
2020	Weech, Wall, Barnett-Cowan	Experimental Brain Research	Noisy GVS during VR games	Short-lived reduction on a verbal sickness scale for intense content only
2022	Groth, Tauscher, Heesen et al.	IEEE TVCG	360-degree video in an HMD	Reduced discomfort for susceptible users
2022	Pradhan, Galvan-Garza et al.	Aerospace Medicine and Human Performance	VR flight simulator, 20 participants	Lower electrogastrogram stress markers, better flight performance
2026	Wagner, Palmisano, Ranjan, Iwasaki, Kim	Virtual Reality	HMD self-motion display	Stronger vection and presence, no change in cybersickness, lower comfort

Safety and limitations

A 2010 review by Utz and colleagues concluded that GVS is safe when standard procedures are used. A 2011 study by the same group analyzed 255 sessions given to 55 people with stroke and 30 healthy individuals at a mean sub-sensory level of 0.6 mA or at 1.5 mA. The most frequent effects were slight itching (10.2 percent) and tingling (10.7 percent) under the electrodes; adverse effects were more frequent at 1.5 mA, and no seizures, vertigo or nausea were observed.

Results in VR are mixed. The studies above report benefits for sickness, vection or presence, but also higher nausea when GVS is deliberately mismatched, reduced comfort, and effects that fade quickly once stimulation stops. In the macaque recordings of Kwan and colleagues, afferent tuning to GVS differed from tuning to natural self-motion, so a GVS signal is not a direct copy of the input produced by real head movement.

See also

• Vestibular system
• Simulator sickness
• Vection
• Entrim 4D
• Motion simulator
• Presence