A bare foot seen from the side, heel on the floor, toes pressing a blue Toe-Rx trainer down into the carpet.
Grip · the second move

The part that works is the part you do.

Still wearing it, lift your toes, then press them down into the floor, squeezing against the foam as they close. Fifteen to twenty times, each foot. This is the working part.

In the trial that separated the two, the people who added this movement improved in every measure; the people who only wore the device improved in none. This page is about why it is done this way — from a spread position, from lifted toes, and not by curling a towel — and what that has and hasn't been shown to change.

The foot has a core ↓ ← Move one: Open
The foot has a core

Two sets of muscles move your toes. Only one of them lives in your foot.

The muscles that curl your toes down are mostly not in your foot at all. They sit in the calf and reach the toes by long tendons that run behind the inside of the ankle. The foot has its own muscles too — four layers of them, small, entirely inside the sole — and in 2015 a group of sports-medicine researchers proposed calling them what they are: the foot's core.

Two-panel anatomical illustration. Left, the sole of the foot seen from below with the intrinsic muscles drawn in red and labelled 'inside the foot'. Right, the inside of the lower leg and foot with the long toe flexors drawn in purple running from the calf, behind the ankle, to the toes, labelled 'from the calf'.
Red: the muscles inside the foot. Purple: the long flexors that reach the toes from the calf. This colour code runs through the rest of the page.
Inside the foot — intrinsic

Small, short, and built to hold the arch.

The abductor hallucis, the short flexors, the lumbricals and interossei. They span the arch from underneath and control how far it flattens under load and how it springs back. They are to the foot what the deep trunk muscles are to the spine — not the movers, the stabilisers.

From the calf — extrinsic

Long, powerful, and aimed at the toe tips.

The long flexors of the toes and big toe. They originate on the shin bones, cross the ankle, and pull the toes down hard. They are the muscles you feel when you curl your toes. They are not the ones that hold your arch up.

Why this distinction matters: it decides what an exercise trains. Measured with fine-wire electrodes and ultrasound, the intrinsic muscles are the ones that respond to load on the arch — activating more as the foot is pressed harder, and storing and returning energy as the arch flattens and recoils. Train the intrinsics and you train the arch's own support. Train the extrinsics and you train a curl. The next section is about how most foot exercises get that wrong.

What "core" does and doesn't mean. It is a framework the authors proposed for thinking about the foot, not a diagnosis or a finding. Its point is that the small muscles were neglected in rehabilitation and deserve the same attention the trunk gets. It does not say that training them fixes any condition — the trials that tested what training them does are further down this page, with their limits.

McKeon PO, Hertel J, Bramble D, Davis I. The foot core system: a new paradigm for understanding intrinsic foot muscle function. British Journal of Sports Medicine, 2015;49(5):290. · Kelly LA, Cresswell AG, Racinais S, Whiteley R, Lichtwark G. Intrinsic foot muscles have the capacity to control deformation of the longitudinal arch. Journal of the Royal Society Interface, 2014;11(93):20131188. · Kelly LA, Farris DJ, Cresswell AG, Lichtwark GA. Intrinsic foot muscles contribute to elastic energy storage and return in the human foot. Journal of Applied Physiology, 2019;126(1):231–238.

Why towel curls miss

The usual foot exercise trains your calf.

Ask almost anyone for a foot-strengthening exercise and you will get the towel curl: put a towel on the floor, scrunch it toward you with your toes. It feels like it works the foot. Measured with electrodes, it mostly works the calf. Curling the toes down is what the long flexors do, and they do it from the lower leg.

Two-panel illustration, same inside view of a lower leg and foot. Left, 'Towel curl': the toes curl down to scrunch a grey towel; the long toe flexors in the calf and their tendons are lit bright purple, the small muscles under the arch are faint. Right, 'Toe-Rx grip': the toes are held spread in a blue Toe-Rx trainer and press down into the floor; the small muscles under the arch are lit bright red, the calf muscles are faint.
Same leg, same view. The colour shows which muscles are doing the work: purple from the calf, red inside the foot.
Measured — 2011

Short-foot exercise lit the big toe's abductor four times harder than the toe curl.

Twenty people, electrodes on the abductor hallucis, standing on one leg. The short-foot exercise — drawing the ball of the foot toward the heel without curling the toes — produced 73% of the muscle's maximum. The toe curl produced 17%. The short-foot exercise also reduced the arch angle; the curl did not.

Measured — 2026

The big toe's long flexor — a calf muscle — was busiest during the towel curl.

Seventeen women with hallux valgus performed nine foot exercises with electrodes on both muscle groups. The long flexor of the big toe, which runs from the calf, was most active during the towel curl and least active during the seated short-foot and the toe-spread-out — the two exercises that keep the toes long.

Abductor hallucis activation, single-leg stance (2011 study)

Short-foot exercise
73%
Toe curl
17%

Percent of maximum voluntary contraction. Jung et al., 2011.

The pattern is the same from both directions. Keep the toes long and work the foot, and the muscles inside the foot light up. Curl the toes, and the calf takes over. The towel curl is not a bad exercise for the calf. It is just not the exercise people think it is. And it is the exercise almost everyone is given.

What we do not say. That towel curls cause deformities, shin splints, or anything else — the old version of this page said so, and nothing supports it. The measurements show which muscles an exercise recruits. They do not show harm from the wrong one. If you like towel curls, keep them; just know what they train.

Jung DY, Kim MH, Koh EK, Kwon OY, Cynn HS, Lee WH. A comparison in the muscle activity of the abductor hallucis and the medial longitudinal arch angle during toe curl and short foot exercises. Physical Therapy in Sport, 2011;12(1):30–35. PMID 21256447. Read it · Electromyographic comparison of nine foot exercises in women with hallux valgus, 2026. PMID 42129706. Read it — full citation to be added.

Why from wide

Not curled. Lifted, spread, and pressing down.

If the towel curl trains the calf, what trains the foot? The previous section gave one answer: keep the toes long. Toe-Rx adds two more conditions — hold the toes apart, and press from lifted — and each has a reason behind it.

Three-panel illustration of the same foot from the inside. Left, 'Curled — not this': the toes curled tightly down, the calf flexors lit purple. Middle, 'Neutral — the shoe position': the toes flat and pressed together, no muscles lit. Right, 'Lifted, wide, pressing — this': the toes spread in a blue Toe-Rx trainer and pressing straight down, a downward arrow at their base, the small muscles under the arch lit red.
The third position is the one the trial and the football study picked out. The first is the one most exercises use. The second is where your toes spend the day.

Two reasons for the third position

Reason one — wide

With the toes held apart, the big toe's abductor is already engaged.

The toe-spread-out exercise was designed for mild hallux valgus: pull the big toe outward against something, so the muscle that abducts it works against the muscle that adducts it. Measured with electrodes, it recruits the abductor hallucis in a way the short-foot exercise alone does not target.

Closing the toes from a spread position works that pair together. Curling from neutral does not — the big toe simply flexes, and the abductor never enters it.

Reason two — lifted

Pressing from lifted toes is the position that mattered in athletes.

In seventeen college football players, toe push-off force with the toes lifted correlated with change-of-direction ability. Toe strength with the toes curled did not correlate with anything. And when toe flexors were trained hard for seven weeks, the gains showed up in tasks where the toe joints are lifted and loaded — horizontal jump — not in walking or running.

Lift, then press, is that position. The curl is the other one.

And the reason it is done from wide at all

Two-column comparison. Left: 'Wore it — 8 weeks, no extra movement' — big-toe angle, active abduction and abductor muscle size all 'no change'. Right: 'Wore it and moved — 8 weeks, with toe movement exercises' — all three 'improved'.

Twenty-four people, eight weeks. Half wore a toe orthosis; half wore it and pulled the big toe outward against it. The first group changed in nothing. The second improved in the angle, the range, and the size of the abductor hallucis. The trial that grew the muscle held the toes apart and had people work from there. That is the Open step and the Grip step, in one study. The Open half is explained on the other page.

Where the reasoning stops. Nobody has put a football player, or anyone else, in a Toe-Rx and measured what the grip step does to their cutting, their arch, or their abductor hallucis. The three findings above — abduction recruits the abductor, the lifted position is the one that correlated, the trial that worked did both — point the same way. They are the reason the grip is done from wide and from lifted. They are not a test of it, and the next two sections say what the tests that do exist found.

Kim MH, Kwon OY, Kim SH, Jung DY. Comparison of muscle activities of abductor hallucis and adductor hallucis between the short foot and toe-spread-out exercises in subjects with mild hallux valgus. Journal of Back and Musculoskeletal Rehabilitation, 2013;26(2):163–168. · Kim MH, et al. Journal of Physical Therapy Science, 2015;27(4):1019–1022 (PMC4433967). · Yuasa Y, Kurihara T, Isaka T. Journal of Human Kinetics, 2018;64:47–55 (PMID 30429898). · Goldmann JP, et al. Journal of Sports Sciences, 2013;31(4):424–433 (PMID 23106289).

What it changes — and what it doesn't

The trials of foot training, in eight tiles.

Everything above is the reasoning. This is what happens when people actually train the muscles inside the foot for weeks, and someone measures. Four things moved. Four did not. The second row is as important as the first, and most pages that sell foot training leave it out.

Changed

Foot strength Meta-analysis, 13 randomised trials
Balance Same meta-analysis
Arch drop under load Meta-analysis; runners, 8 weeks
Cutting & horizontal jump Athletes, randomised, 8 weeks

And in one year-long trial, runners who did foot-core training were 2.4 times less likely to sustain a running injury than a group doing static stretching. That is the strongest single result in this field, and it is in one sport, in one trial.

Did not change

Pain Not significantly improved, same meta-analysis
Impact force Unchanged; runners, 8 weeks
Sprint speed Unchanged; athletes, randomised
Vertical jump Inconsistent across four trials

Running economy did not change either when toe flexors were trained alone. Nothing in this row is a failure of the exercise. It is the shape of what the exercise does: the foot's own muscles, the arch, and the movements where the foot is loaded at an angle — not pain, not impact, not straight-line speed.

  • 13 trials A 2023 meta-analysis pooled every randomised trial of intrinsic foot muscle training it could find. Strength, balance, navicular drop and self-reported function improved. Pain did not.
  • Runners, 8 wk Foot-core training reduced how far the arch moved during the stance phase of running. Impact force and loading rate did not change.
  • Runners, 1 yr 118 recreational runners, randomised. The foot-core group had 2.4 times fewer running-related injuries than the stretching group over twelve months.
  • Athletes, 8 wk 28 highly trained athletes, randomised. High-load forefoot strengthening improved 90° cutting and horizontal jump; sprint acceleration did not change. Effects held through four weeks of detraining.

One more thing that has not been tested: foam.

The athlete trial loaded the toes on a dynamometer-type setup at high tension, twice a week, supervised. The runner trials used a progressive programme of foot exercises with a physiotherapist. Toe-Rx offers the resistance of molded foam, for ten minutes, on your own. Whether that is enough load to produce the same changes is not known. What it gives you is the position — wide, lifted — and something to close against. We say the movement is what counts, and we say the trials used more of it than a foam block provides.

Jaffri AH, Koldenhoven R, Saliba S, Hertel J. Journal of Athletic Training, 2023;58(11–12):941–951 (PMC10784881). · Matias AB, et al. Frontiers in Bioengineering and Biotechnology, 2022 (doi:10.3389/fbioe.2022.890428). · Taddei UT, Matias AB, Duarte M, Sacco ICN. American Journal of Sports Medicine, 2020 (doi:10.1177/0363546520969205). · Tourillon R, Fourchet F, Edouard P, Morin JB. PLOS ONE, 2025;20(6):e0313979 (PMC12129235). · Goldmann JP, et al. Journal of Sports Sciences, 2013;31(4):424–433. · Day EM, Hahn ME. Journal of Sports Sciences, 2020 (toe flexor training and running economy).

How to grip

Lift. Press. Release. Fifteen to twenty times.

With Toe-Rx already in — after the Open and Circle steps on the other page — this is one repetition. Sitting down, heel on the floor, one foot at a time. The toes stay long the whole way through. If they curl, you have switched to the calf; start the rep again.

Side view of a bare foot with a blue Toe-Rx trainer on the toes. The toes are lifted at their base joints, leaving a gap between the toes and the floor.
1

Lift

Raise all five toes from the joints at their base, device and all, until there is daylight under them. Keep them straight — a lifted board, not a curl. This is the position the football study measured.

Side view of the same foot. The toes are pressing the Toe-Rx trainer down into the floor; the foam is visibly compressed and the toes are straight, pushing from their base.
2

Press

Drive the toes straight down into the floor and squeeze them together against the foam as they go. Hold for a second. You should feel it under the arch, not in the calf. The foam compressing is the resistance.

Side view of the same foot at rest. The toes are relaxed and still spread by the Toe-Rx trainer, resting lightly on the floor.
3

Release

Let go. The foam springs the toes back apart and they rest. That is one. Rest a breath, then lift again.

Fifteen to twenty repetitions, then the other foot. Two to three minutes per side. With the Open and Circle steps before it, the whole routine is about ten minutes per foot. Every other day is enough — the trials ran two sessions a week, and nothing we have read says daily beats that. Full instructions with photos

Best

After activity

Shoes are already off and the day's work is done. If you only ever do it at one time, do it here.

Best

On a rest day

Nothing to work around. The trials delivered it as its own session, not attached to a warm-up.

Not then

Before running, jumping or sport

A static hold of 60 seconds or more has been shown to reduce maximal force for a while afterwards. The Open step is a long hold. Do the routine after, not before.

Where it stops

  • Pain is a clinician's question, not ours. The one thing the meta-analysis did not find improved was pain. If your foot hurts, that is the number to hold on to, and a reason to see someone rather than train through it.
  • Nothing here has been tested on Toe-Rx itself. The trials used other programmes and, in the one that improved cutting, far more load than foam provides. We cite them for the movement, not the device.
  • Not before sport. See above. The minutes before you need to move fast are the wrong minutes for a long hold.
  • Not on an injured toe joint without clearance. A sprained big toe, a healing fracture, a joint that is thick from an old injury — whoever is managing it decides whether anything goes between the toes, not a web page.

The position this starts from — and why holding the toes open is the setup, not the fix — is on the other page: ← Open: why the toes are held apart

References

Every figure on this page comes from one of these. None of them tested Toe-Rx; where a study's findings run out, the section that cites it says so.

  1. McKeon PO, Hertel J, Bramble D, Davis I. The foot core system: a new paradigm for understanding intrinsic foot muscle function. British Journal of Sports Medicine, 2015;49(5):290.— the framework: intrinsic muscles as the foot's core.
  2. Kelly LA, Cresswell AG, Racinais S, Whiteley R, Lichtwark G. Intrinsic foot muscles have the capacity to control deformation of the longitudinal arch. Journal of the Royal Society Interface, 2014;11(93):20131188.— intrinsics activate with load on the arch.
  3. Kelly LA, Farris DJ, Cresswell AG, Lichtwark GA. Intrinsic foot muscles contribute to elastic energy storage and return in the human foot. Journal of Applied Physiology, 2019;126(1):231–238.— intrinsics store and return energy as the arch flattens and recoils.
  4. Jung DY, Kim MH, Koh EK, Kwon OY, Cynn HS, Lee WH. A comparison in the muscle activity of the abductor hallucis and the medial longitudinal arch angle during toe curl and short foot exercises. Physical Therapy in Sport, 2011;12(1):30–35. PMID 21256447— 73% vs 17% abductor hallucis activation, short-foot vs toe curl.
  5. Electromyographic comparison of nine foot exercises in women with hallux valgus, 2026. PMID 42129706— the long flexor of the big toe was most active in the towel curl. Full citation to be added.
  6. Kim MH, Kwon OY, Kim SH, Jung DY. Comparison of muscle activities of abductor hallucis and adductor hallucis between the short foot and toe-spread-out exercises in subjects with mild hallux valgus. Journal of Back and Musculoskeletal Rehabilitation, 2013;26(2):163–168.— the toe-spread-out exercise and the abductor–adductor pair.
  7. Kim MH, Yi CH, Weon JH, Cynn HS, Jung DY, Kwon OY. Effect of toe-spread-out exercise on hallux valgus angle and cross-sectional area of abductor hallucis muscle in subjects with hallux valgus. Journal of Physical Therapy Science, 2015;27(4):1019–1022. doi:10.1589/jpts.27.1019— orthosis alone changed nothing; orthosis plus movement improved all three measures.
  8. Yuasa Y, Kurihara T, Isaka T. Relationship between toe muscular strength and the ability to change direction in athletes. Journal of Human Kinetics, 2018;64:47–55. PMID 30429898— lifted-toe push-off correlated with agility; curled-toe strength did not.
  9. Goldmann JP, Sanno M, Willwacher S, Heinrich K, Brüggemann GP. The potential of toe flexor muscles to enhance performance. Journal of Sports Sciences, 2013;31(4):424–433. PMID 23106289— seven weeks of toe flexor training: horizontal jump improved; vertical, walking, running unchanged.
  10. Jaffri AH, Koldenhoven R, Saliba S, Hertel J. Evidence for intrinsic foot muscle training in improving foot function: a systematic review and meta-analysis. Journal of Athletic Training, 2023;58(11–12):941–951. Read it— 13 trials: strength, balance, navicular drop and function improved; pain did not.
  11. Matias AB, Watari R, Taddei UT, et al. Effects of foot-core training on foot-ankle kinematics and running kinetics in runners. Frontiers in Bioengineering and Biotechnology, 2022. doi:10.3389/fbioe.2022.890428— arch excursion reduced; impact force and loading rate unchanged.
  12. Taddei UT, Matias AB, Duarte M, Sacco ICN. Foot core training to prevent running-related injuries: a survival analysis of a single-blind, randomized controlled trial. American Journal of Sports Medicine, 2020. doi:10.1177/0363546520969205— 2.4× fewer running injuries over one year.
  13. Tourillon R, Fourchet F, Edouard P, Morin JB. Effects of a forefoot strengthening protocol on explosive tasks performance and propulsion kinetics in athletes: a single-blind randomized controlled trial. PLOS ONE, 2025;20(6):e0313979. Read it— cutting and horizontal jump improved; sprint acceleration did not.
  14. Day EM, Hahn ME. Does toe flexor training affect running economy? Journal of Sports Sciences, 2020.— running economy unchanged. Volume and pages to be added.
  15. Revisiting the stretch-induced force deficit: a systematic review with multilevel meta-analysis of acute effects. Journal of Sport and Health Science, 2024. Read it— the 60-second threshold behind the timing advice.

These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease.