Gaze during twists and pirouettes in gymnastics, trampoline, dance and skating: what 30 studies describe

In brief. In a somersault, twist or pirouette, the expert places their gaze on a landmark at the start and before landing; in the heart of a fast twist, their eyes stop stabilizing the image. Twenty-nine of these 30 studies observe or test athletes without training them; the only training trial, spotting taught to novice dancers, did not improve their pirouettes. Updated on November 3, 2026.

Hello to you, the movement professional,

In Montreal, seventeen trampolinists, eight elite and nine sub-elite, performed four twisting somersaults of increasing difficulty with an eye tracker and seventeen inertial measurement units on the body (Charbonneau et al., 2024). Besides the expected visual landmark at the start and the end of the acrobatic skill, the recordings show what the authors call "self-motion detection": when the head turns fast, especially during the twist, the eyes no longer move in the orbit.

Of the 30 studies in this file, twenty-nine train no one: they observe, deprive of vision for the length of a session, question experts or simulate. The thirtieth taught spotting to novice dancers for eight weeks, with no better pirouette to show for it. You will read descriptions, not a training method.

Why would an expert give up stabilizing their gaze at the moment when everything spins fastest?

Gaze during a gymnastics twist or a pirouette: what the eyes and the head do

The vestibulo-ocular reflex (VOR) turns the eyes in the opposite direction to the head to keep the image still on the retina. The semicircular canals of the inner ear pick up the rotation, and the vestibular nuclei of the brainstem relay the signal to the nuclei that command the eye muscles. When the head turns too fast, the image slips and blurs, which is what dynamic visual acuity measures in the ball-sport athlete.

In twelve experienced female acrobats, liquid crystal glasses cut vision depending on head speed, above or below 350 degrees per second (Luis and Tremblay, 2008). Any vision was better than none for the landing, and seeing only during the slow phases even ranked ahead of full vision. According to the authors, visual information is useful at all speeds, and best when the image stays stable on the retina.

Dancers call their way of turning spotting: the head turns apart from the trunk to keep a visual landmark. Experts consulted with the Delphi method grant it head isolation, timing and gaze precision, with weak consensus (Haber and Schärli, 2021). In these studies, the expert's gaze works mostly at the start and before landing, when the head turns less fast.

Expert gymnasts and trampolinists: fixating the bed, not blinking, fixating later

In a backward somersault on the trampoline, ten experts land closer to the center of the bed than ten apprentices, fixate more at takeoff and in flight, and never blink (Heinen, 2011a). On a straight jump with a full turn, ten gymnasts stabilize their gaze before takeoff, then about 110 milliseconds before landing, earlier than fourteen non-gymnasts (Sato et al., 2017).

In Münster, eleven trampolinists all fixate a point on the bed during the tucked backward somersault, and their eyes move down continuously to counter the head going backward (Natrup et al., 2020). On the straight backward somersault with a full twist, fifteen athletes move their eyes with the head in the direction opposite to the twist (Natrup et al., 2021). In both studies, the best category fixates the bed later, perhaps in favor of form according to the authors; on the twist, this gap is significant. The less qualified seem to look for their landmark as early as possible, with the head more turned and a poorer execution, but this is only a non-significant trend.

In Montreal, the elites make more fixations, the only expertise difference out of six measures (Charbonneau et al., 2024), and the variability of pelvis orientation drops by 39% before landing, especially when the athlete looks at the bed, a link observed without a statistical test (Bourgeois et al., 2025). On the run-up, the vaulting table changes gaze only in elites, not in super-elites (Barreto et al., 2021).

What remains of the landing when vision is removed?

Somersault without the eyes: what vision brings to the landing

In trampolinists landing from a forward somersault, body extension is tuned to the time remaining before contact, through a perceptual variable called tau, and vision makes this tuning more precise (Lee et al., 1992). In ten young gymnasts, reducing the peripheral field to 100 then 60 degrees changes nothing, whereas removing all vision destabilizes the landing (Davlin et al., 2001a); seeing during the second half of the somersault is enough (Davlin et al., 2001b). In a double backward somersault, university acrobats keep their scores with central vision blocked, lose them without vision, and slow the head before contact to pick up the visual flow (Hondzinski and Darling, 2001).

The sport practiced matters: without vision, eleven parkour traceurs degrade on the backward somersault, seven gymnasts do not (Mohamed Abahnini et al., 2023). On the vault, removing binocular vision leaves the experts unchanged and modifies gaze and vault phases in the apprentices (Heinen and Vinken, 2011b). In eighteen gymnasts of about 14, a gaze instruction, downward or ahead, has little influence on stabilization at landing (Pavlasová et al., 2026). In simulation, visual criteria bring the computed techniques closer to that of an elite athlete (Charbonneau et al., 2026).

During the fast twist, the eyes follow the body

A phenomenon close to the one seen in Montreal was described ten years earlier: in high-level gymnasts in somersaults with multiple twists, von Laßberg and colleagues observe a prolonged and complete suppression of the movements that stabilize gaze, replaced by a tight coupling of the eyes with the head and the spine (von Laßberg et al., 2014). Before landing, stabilization takes its place back, the team's next study specifies.

A VOR adapted by dint of spinning? This 2020 study rather answers no: followed for three years in a vestibular laboratory, the VOR of high-level gymnasts varies a great deal from one person to another without differing systematically from that of controls (von Laßberg et al., 2020). The authors lean toward eyes driven by reflexes of proprioceptive origin, even by the motor plan. During the fastest phase, the expert lets go of the image of the world and glues their eyes to the movement of their body. This reading remains a hypothesis of the authors, which their VOR measurements make plausible without demonstrating it.

Spotting in dance: less dizziness in the expert, no pirouette gain in the novice

In Bern, twelve high-level ballet dancers performed twelve fouettés or turns à la seconde, with and without a visual landmark (Schärli et al., 2023). Across all the turns, the path of the supporting foot is even shorter without a landmark; over the course of the series, however, the landmark shortens this path and improves orientation: it would mainly help sustain a long series, according to the authors.

The same team had thirty-four dancers turn, fourteen standing turns or fourteen passive rotations on a motorized chair, with and without spotting (Schärli et al., 2024). Without spotting, dizziness is stronger in both cases; balance measured on a force plate benefits only after the passive rotations.

That left passing the skill on. For eight weeks, novice dancers received spotting instructions, others technical instructions (Klostermann et al., 2022). At retention, few still used spotting, their double pirouettes were no better, and their balance stagnated while that of the controls improved; only orientation kept a slight advantage. The technique "is not suited" to training, "at least for beginners," the authors conclude (our translation).

Dancers, skaters and gymnasts: a vestibulo-ocular reflex tuned differently

After a rotation, the eyes still beat for several seconds (post-rotatory nystagmus). In twelve ballet dancers, braking it through vision works better than in untrained people, and all the better the longer the experience (Teramoto et al., 1994). On the video head impulse test (vHIT), twelve dancers suspend their VOR better than twelve controls to follow a head-fixed target; the authors point to the pathway between cerebellum and vestibular nuclei (Maheu et al., 2019). In thirteen gymnasts, this suppression is better in the most experienced, the VOR remaining comparable at all levels (van der Veen et al., 2022).

On VOR gain, results diverge: Maheu's team finds it higher in dancers, on the left only (Moin-Darbari et al., 2023). In nine female dancers, it stays within norms, with no significant difference from nine female non-dancers (Jha et al., 2025): in this small sample, nothing suggests an inner ear "put to sleep" by dance. In eleven female figure skaters, it is 27 to 32% lower than in eleven controls, with a much weaker induced motion sickness, a probable sign of habituation according to the authors (Tanguy et al., 2008a); on this sickness, see the link between visual field and kinetosis. In twelve skaters, the signal from the otoliths seems to be read more as a rotation than as a translation (Tanguy et al., 2008b).

Smooth pursuit also differs: higher gain in elite gymnasts than in young national-team members, and in them than in non-athletes (von Laßberg et al., 2012). In the batter who follows the ball with the eyes, this pursuit works in a completely different context; our article on keeping gaze stable when everything moves details the reflex.

Links between the studies

Several works share a team, sometimes probably a sample: the two Davlin studies describe ten gymnasts with an identical profile, Charbonneau 2024 and Bourgeois 2025 seventeen trampolinists from the same Montreal laboratory, and the two Tanguy studies, in Caen, eleven then twelve skaters who may overlap. Maheu and Moin-Darbari work in Montreal on dancers, Schärli, Haber and Klostermann in Bern: treat these results as less independent than they look.

What these 30 studies do not yet allow us to claim

Established, on small samples (ten to thirty-four participants per study) but by several teams: the rotation expert places their gaze on a landmark at the start and before landing, and removing all vision most often destabilizes the landing (seven gymnasts escaped this in one study, compared with parkour traceurs). Debated: the fate of the VOR in those who spin, and a better suppression seen only in cross-sectional comparison. Extrapolated: that teaching an expert's gaze strategy speeds up the learning of a rotation; the only trial, in the novice, finds no benefit for the pirouette. It remains to be seen whether these differences come from practice or from a selection of those who tolerate spinning best.

One prediction can settle it: if the dancers' VOR suppression comes from practice, a follow-up of beginners over several years should see it appear; if it is there from the start, selection explains it.

Before touching the gaze of an athlete who spins: the ophthalmologist first

If your athlete reports double vision, a blur that persists, headaches during effort or dizziness, visual training waits. They first see an ophthalmologist, then an orthoptist in France, or an optometrist in Switzerland or the United States. LabO RNP is neither medical nor paramedical, and comes in addition to a care pathway, never in its place. A slight passing dizziness after a series of turns is usual; a dizziness that persists at rest is shown first to a doctor, before any visual or rotational work.

Seeing to land: vision in the sensorimotor loop

The eyes are an entry point; LabO RNP teaches the whole sensorimotor loop, where what the athlete perceives organizes what they do. In a twist, inner ear, neck and eyes share the work, and the expert chooses when to see. For the place of the eyes in the warm-up, read our oculomotor exercises for the warm-up; for all the families of visual training, the map of the evidence in sports vision; for the other half of the story, our articles on the vestibular sensory sensors.

Gaze and rotations: the questions that come up in the gym

Where does a gymnast look during a somersault with a twist?

On the trampoline, they fixate the bed before landing, eyes and head moving together opposite to the twist, and the best athletes fixate later than the others (Natrup et al., 2021). At the fastest point of the rotation, their eyes stop stabilizing the image and follow the movement of the body (von Laßberg et al., 2014; Charbonneau et al., 2024). Gaze lands mostly at the start and before landing.

Does spotting prevent dizziness in dance?

It reduces it. In Bern, thirty-four ballet or contemporary dancers rated their dizziness on a scale of 0 to 20 after fourteen standing turns and after fourteen rotations on a motorized chair: it was stronger without spotting in both cases. Balance measured on a force plate benefited only after the passive rotations (Schärli et al., 2024).

Should you teach spotting to a beginner dancer?

The only trial does not support it. After eight weeks of spotting instructions, novice dancers did not perform the double pirouette any better than those who had received technical instructions, few kept the spotting, and their balance stagnated while that of the controls improved (Klostermann et al., 2022). In high-level dancers, a visual landmark mainly helps sustain a long series of turns (Schärli et al., 2023).

Is working on a gymnast's gaze a form of treatment?

No. LabO RNP is neither medical nor paramedical, and comes in addition to a care pathway, never in its place. Double vision, a blur that persists, headaches during effort or dizziness: the ophthalmologist first, then the orthoptist in France or the optometrist in Switzerland and the United States.

The expert's gaze knows when to see, and it is the landing that tells whether they saw in time.

👉 I want to understand what my athlete perceives when they spin.

By the LabO RNP team

Method. LabO RNP inventory of the athlete's vision, completed in PubMed on November 3, 2026: 37 studies identified on gaze in rotation, 30 analyzed, including Klostermann 2022, found outside the inventory. Excluded: seven references without abstract, without PMID, without results or outside rotation, as well as the gaze of judges and the static balance of dancers. PMIDs verified on the same date.

Sources

• Barreto J, Casanova F, Peixoto C, Fawver B, Williams AM (2021). How Task Constraints Influence the Gaze and Motor Behaviours of Elite-Level Gymnasts. Int J Environ Res Public Health, 18(13). PMID 34209486. doi:10.3390/ijerph18136941.

• Bourgeois M, Charbonneau E, Turner C, Begon M (2025). Motor variability regulation analysis in trampolinists. J Biomech, 181, 112533. PMID 39862717. doi:10.1016/j.jbiomech.2025.112533.

• Charbonneau E, Begon M, Romeas T (2024). A temporal quantitative analysis of visuomotor behavior during four twisting somersaults in elite and sub-elite trampolinists. Hum Mov Sci, 98, 103295. PMID 39378631. doi:10.1016/j.humov.2024.103295.

• Charbonneau E, Romeas T, Ross A, Begon M (2026). Including visual criteria into predictive simulation of acrobatics to enhance the realism of optimal techniques. Sports Biomech, 25(5), 726-748. PMID 41222025. doi:10.1080/14763141.2025.2577924.

• Davlin CD, Sands WA, Shultz BB (2001a). Peripheral vision and back tuck somersaults. Percept Mot Skills, 93(2), 465-71. PMID 11769904. doi:10.2466/pms.2001.93.2.465.

• Davlin CD, Sands WA, Shultz BB (2001b). The role of vision in control of orientation in a back tuck somersault. Motor Control, 5(4), 337-46. PMID 11581503. doi:10.1123/mcj.5.4.337.

• Haber C, Schärli A (2021). Defining Spotting in Dance: A Delphi Method Study Evaluating Expert Opinions. Front Psychol, 12, 540396. PMID 34054627. doi:10.3389/fpsyg.2021.540396.

• Heinen T (2011a). Evidence for the spotting hypothesis in gymnasts. Motor Control, 15(2), 267-84. PMID 21628729. doi:10.1123/mcj.15.2.267.

• Heinen T, M Vinken P (2011b). Monocular and binocular vision in the performance of a complex skill. J Sports Sci Med, 10(3), 520-7. PMID 24150627. No DOI assigned. PMC3737810.

• Hondzinski JM, Darling WG (2001). Aerial somersault performance under three visual conditions. Motor Control, 5(3), 281-300. PMID 11463047. doi:10.1123/mcj.5.3.281.

• Jha RH, Piker EG, Scalzo M, Trinidad D (2025). Dance Training and the Neuroplasticity of the Vestibular-Ocular Reflex: Preliminary Findings. Brain Sci, 15(4). PMID 40309842. doi:10.3390/brainsci15040355.

• Klostermann A, Schärli A, Kunz S, Weber M, Hossner EJ (2022). Learn to Turn: Does Spotting Foster Skill Acquisition in Pirouettes?. Res Q Exerc Sport, 93(1), 153-161. PMID 32931382. doi:10.1080/02701367.2020.1813239.

• Lee DN, Young DS, Rewt D (1992). How do somersaulters land on their feet?. J Exp Psychol Hum Percept Perform, 18(4), 1195-202. PMID 1431752. doi:10.1037//0096-1523.18.4.1195.

• Luis M, Tremblay L (2008). Visual feedback use during a back tuck somersault: evidence for optimal visual feedback utilization. Motor Control, 12(3), 210-8. PMID 18698106. doi:10.1123/mcj.12.3.210.

• Maheu M, Behtani L, Nooristani M, Delcenserie A, Champoux F (2019). Enhanced vestibulo-ocular reflex suppression in dancers during passive high-velocity head impulses. Exp Brain Res, 237(2), 411-416. PMID 30426147. doi:10.1007/s00221-018-5431-z.

• Mohamed Abahnini N, Abahnini K, Mkaouer B (2023). Effects of eliminating visual cues on kinetic and kinematic parameters in back tuck somersault: A comparison between artistic gymnasts and parkour athletes. Acta Bioeng Biomech, 25(2), 93-101. PMID 38314533. No DOI assigned.

• Moin-Darbari K, Nooristani M, Bacon BA, Champoux F, Maheu M (2023). Long-term dance training modifies eye-head coordination in response to passive head impulse. J Neurophysiol, 130(4), 999-1007. PMID 37702547. doi:10.1152/jn.00232.2023.

• Natrup J, Bramme J, de Lussanet MHE, Boström KJ, Lappe M, Wagner H (2020). Gaze behavior of trampoline gymnasts during a back tuck somersault. Hum Mov Sci, 70, 102589. PMID 32217208. doi:10.1016/j.humov.2020.102589.

• Natrup J, de Lussanet MHE, Boström KJ, Lappe M, Wagner H (2021). Gaze, head and eye movements during somersaults with full twists. Hum Mov Sci, 75, 102740. PMID 33307374. doi:10.1016/j.humov.2020.102740.

• Pavlasová K, Bizovská L, Rupčík L, Farana R, Janura M (2026). The Assessment of the Effect of Gaze Direction Instruction on the Stabilisation During Artistic Gymnastic Landing. Eur J Sport Sci, 26(3), e70137. PMID 41651765. doi:10.1002/ejsc.70137.

• Sato Y, Torii S, Sasaki M, Heinen T (2017). Gaze-Shift Patterns during a Jump with Full Turn in Male Gymnasts. Percept Mot Skills, 124(1), 248-263. PMID 27837042. doi:10.1177/0031512516676148.

• Schärli A, Haber C, Klostermann A (2023). Does a visual reference help ballet dancers turn more successfully?. Hum Mov Sci, 88, 103062. PMID 36680891. doi:10.1016/j.humov.2023.103062.

• Schärli A, Hecht H, Mast FW, Hossner EJ (2024). How spotting technique affects dizziness and postural stability after full-body rotations in dancers. Hum Mov Sci, 95, 103211. PMID 38583276. doi:10.1016/j.humov.2024.103211.

• Tanguy S, Quarck G, Etard O, Gauthier A, Denise P (2008a). Vestibulo-ocular reflex and motion sickness in figure skaters. Eur J Appl Physiol, 104(6), 1031-7. PMID 18758803. doi:10.1007/s00421-008-0859-7.

• Tanguy SG, Quarck GM, Etard OM, Gauthier AF, Denise P (2008b). Are otolithic inputs interpreted better in figure skaters?. Neuroreport, 19(5), 565-8. PMID 18388739. doi:10.1097/WNR.0b013e3282f9427e.

• Teramoto K, Sakata E, Ohtsu K (1994). Use of the visual suppression test using post-rotatory nystagmus to determine skill in ballet dancers. Eur Arch Otorhinolaryngol, 251(4), 218-23. PMID 7917255. doi:10.1007/BF00628427.

• van der Veen SM, Stamenkovic A, Thomas JS, Pidcoe PE (2022). Skill-Related Adaptive Modifications of Gaze Stabilization in Elite and Non-Elite Athletes. Front Sports Act Living, 4, 824990. PMID 35498518. doi:10.3389/fspor.2022.824990.

• von Laßberg C, Beykirch KA, Mohler BJ, Bülthoff HH (2014). Intersegmental eye-head-body interactions during complex whole body movements. PLoS One, 9(4), e95450. PMID 24763143. doi:10.1371/journal.pone.0095450. Erratum published in PLoS One, October 2014.

• von Laßberg C, Campos JL, Beykirch KA (2020). Long term measures of vestibulo-ocular reflex function in high level male gymnasts and its possible role during context specific rotational tasks. PLoS One, 15(12), e0243752. PMID 33315913. doi:10.1371/journal.pone.0243752.

• von Lassberg C, Beykirch K, Campos JL, Krug J (2012). Smooth pursuit eye movement adaptation in high level gymnasts. Motor Control, 16(2), 176-94. PMID 22388227. doi:10.1123/mcj.16.2.176.