Tack and equipment: what the evidence says about the gear on your horse

You finish a ride, swing the saddle off, and notice the patch of damp hair behind your horse's withers sits a little further forward than it used to. The noseband leaves a faint imprint. Your trainer mentions a new bit. You wonder, quietly, whether any of the gear is doing what the catalogue said it would, and whether any of it is quietly working against your horse.

That is the question this guide takes seriously. Most equestrian publishing is shaped by the same brands the research is testing, so owners end up reading marketing rather than methods. The good news is that the past three years of tack research have produced unusually clean answers on a handful of points. Some of those answers will tell you to keep doing what you are doing. Some will tell you the opposite.

A note on language. When studies show "no evidence" for a piece of gear, that means the trial has not been done. When studies show "evidence of no effect", the trial was done and the product failed. Those are very different verdicts, and conflating them is how owners pay for things that have never been tested.

Is riding your horse in deep flexion ever worth it?

Evidence of no effect

Welfare costs of hyperflexion are documented across the literature. Performance benefits are inconsistent and on balance null.

Hyperflexion (the forced low-deep-round head and neck position, sometimes called rollkur) has been one of the most contested practices in dressage. The biggest piece of evidence to date is a 2024 meta-analysis that pulled together fifty-eight peer-reviewed hyperflexion studies. Three-quarters of those studies raised welfare concerns. Performance effects split twenty-five percent advantage, twenty-five percent detriment, and forty-four percent null. The authors concluded that the welfare cost outweighs any performance benefit.

Experimental work backs that up. A 2023 study tested forty elite dressage and show-jumping horses at one hundred versus eighty-five degrees of poll flexion. That fifteen-degree difference produced more conflict behaviours and more upper airway abnormalities at the deeper angle, and more relaxation behaviours at the shallower one. A 2025 competition study coded conflict behaviour across one hundred and thirty-five dressage rounds and found mouth opening in every single horse and tail swishing in nearly nine out of ten, regardless of how the tack was set up.

A 2025 governance review places hyperflexion alongside double bridles and tight nosebands as practices where rule-making has fallen behind what the evidence shows.

How tight is too tight on a noseband?

Worth it

The two-finger rule is supported by convergent evidence. Pressures rise rapidly below 1.4 fingers; stride length, airway function, and ridden behaviour all degrade as tightness increases.

The cleanest dose-response story in this whole pillar is on noseband tightness. If you only check one thing on your horse before a ride, this is it.

A 2025 cadaver-head study (cadaver-head means research on a horse's preserved head, which lets researchers measure pressure without behaviour getting in the way) put pressure sensors on the nasal bone, the frontal plane, and inside the mouth at the second premolar, then tightened the noseband in eight steps using a standard taper gauge. Peak pressure hit four hundred and three kilopascals at the nasal bone when the noseband was at zero fingers, and the rise got steep once tightness passed roughly one and a half fingers.

Live horses tell the same story. In a 2025 ridden study, eight horses were ridden at two-finger, one-finger, and zero-finger settings. Pressure rose fifty-four percent at one finger and three hundred and thirty-eight percent at zero fingers compared to two fingers. Stride length dropped six and eleven percent at the same settings. A trot study replicated the pattern across cavesson, Swedish, drop, and flash designs.

Inside the airway, the picture is similar. A 2024 endoscopy study rode sixteen Warmbloods on consecutive days under loose and tight nosebands while a camera looked at pharyngeal function. Tight nosebands produced more secretion, more pharyngeal collapse, and higher RHpE scores (the ridden horse pain ethogram, a checklist of twenty-four behaviours that signal ridden pain).

There is an active methodological argument about whether checking tightness against the side of the jaw underestimates how tight the band really is. The fix is straightforward. Check on the dorsal midline, the bony bridge of the nose, with two fingers stacked flat between leather and bone. If they do not slide easily, the noseband is too tight.

Can the ridden horse pain ethogram tell you what is wrong?

Mixed evidence

The Ridden Horse Pain Ethogram is a useful screen for ridden discomfort. It is not a diagnostic. Gastric disease can elevate scores.

The RHpE has become a popular tool, and the research on it is genuinely encouraging, with caveats. In a 2023 study, one hundred and fifty horses with musculoskeletal pain were assessed before and after diagnostic anaesthesia plus saddle adjustment. The median score dropped from nine to two out of twenty-four after intervention, and about thirty-seven percent of cases involved a poorly fitting saddle. A companion paper applied the tool to one thousand three hundred and fifty-eight competition horses and found median scores at four with median lameness at zero, which means pain-related behaviours show up regularly even when the horse is not visibly lame.

Two things keep this from being a clean "worth it." First, most of the strongest applications come from the group that developed the tool, and independent replication is still thin. Second, signal specificity is incomplete. A 2024 study followed nine show-jumping Warmbloods through a diet change from high-sugar pellets to a low-starch muesli and found that gastric disease scores and RHpE scores both improved, with strong correlation. In other words, your horse can score high on a musculoskeletal screening tool because his stomach hurts.

Use the RHpE as a flag that something is worth investigating. Do not use it to diagnose what.

Does saddle fit have to be measured during movement?

Mixed evidence

Dynamic measurement during locomotion adds value over static-only fitting. The mechanistic case is strong.

A standing horse is not the same shape as a moving horse. A 2024 photogrammetry study used multiple cameras on five horses and found backs are flatter while moving than while stationary, with the largest deformation right at the withers. A 2025 IMU study (IMU stands for inertial measurement unit, a small motion sensor) tested eight rider-horse pairs across four saddles that differed only in seat depth, flap width, and front panel thickness. Even minor differences produced detectable changes in how horse and rider moved together.

It gets more complicated. A 2025 girth study found that higher girth tension shifts saddle pressure forward in a way that matters for thoracic back pain risk, regardless of how the girth was designed. Earlier work on thigh blocks found that deformable foam-layered blocks changed both rider posture and horse kinematics compared to rigid vertical blocks. That paper is industry-affiliated, methodology sound, worth a flag.

The owner-facing question is settled by a 2025 survey. Four hundred and one respondents looked at photographs of saddles in place and identified positioning correctly fifty-eight percent of the time. Experience did not improve the score. If you are calibrating against your own eye, you are wrong about four times in ten, and that is true whether you have ridden for one year or twenty.

Is a double bridle harder on a horse than a snaffle?

Mixed evidence

Force redistribution shifts load to the poll and reduces per-rein tension by splitting force across two bits. Behavioural conflict prevalence is high regardless of bridle type, which complicates a clean call.

A 2025 study rode eleven elite dressage horses in a snaffle and then a double bridle. Noseband pressures came out similar across the two. Force on the poll was higher under the double (the hardware is heavier). Tension on each individual rein was lower in the double because the load was split across two bits. The paper is industry-funded, methodology robust.

Behaviourally, the picture is muddier. The 2025 competition data found conflict markers ubiquitous regardless of bridle type. A 2026 paper reported more conflict behaviours in double-bridled horses than snaffle-bridled horses in the same population, though the comparison is observational and tangled with competition level. A 2025 rein tension study looked at heart rate variability and found tension increased sympathetic activity, though welfare conclusions outpace what those shifts can directly establish.

The honest read is that snaffle versus double is not the variable that most determines your horse's experience. Rider hands, contact technique, and noseband configuration carry similar or larger weight.

What about treeless saddles, spurs, and training aids?

Absence of evidence

No peer-reviewed 2023-2026 study tests treeless saddles directly. No 2023-2026 primary study isolated to spur use was identified. Training aid effects on training outcomes are not in the evidence base.

These three categories share a verdict, which is that the marketing claims are running ahead of the trials. The dynamic-back-shape work gives mechanistic reason to think saddles that conform to a moving back could be useful, but mechanism is not outcome, and a controlled comparison of treeless against well-fitted treed saddles does not exist in the modern literature.

Spurs sit in a similar position. They are widely discussed in welfare governance, but the controlled studies isolating spur use are not there at this window. A 2023 lunging-aid study compared free-head, chambon, rubber-band side rein, and triangle side rein configurations in thirteen horses and found postural and surface temperature differences across configurations. Differences in posture are not the same as differences in welfare or training outcomes, and that link has not been studied. A 2026 paper does provide one clean adjacent finding: purpose-made foam padding under a Pessoa training aid roller cuts mean pressure by seventy-five to eighty-two percent compared to no padding.

If a salesperson or trainer makes a strong claim about a treeless saddle, a spur, or a training aid, the honest answer at three years is that the trial supporting the claim has not been published.

What does the whip actually do when you measure it?

Evidence of no effect

When researchers synchronise whip strikes to speed and stress data, the measured response runs opposite to the rider's intent. Strikes are followed by slowing, and stress markers rise with whip frequency.

Whip use is one of those tack questions where the assumed answer (a tap reinforces the leg, a stronger strike sharpens response) has gone largely unmeasured outside the lab. When researchers do measure it, the picture that comes back is not the one most riders would predict.

A 2025 study provides the cleanest behavioural test to date. Researchers coded two hundred and sixty-eight whip strikes across forty-eight horses, with video synchronised to positional speed data. The strikes most often produced deceleration, with statistically significant slowing within three seconds of a strike. Trotters under harness are not ridden sport horses, and the finding does not generalise wholesale, but the design is unusual in that it directly tests the handler's intent (faster) against the measured outcome (slower) and gets the opposite of the assumed answer. For any rider who carries a stick on the basis that escalation produces more forward, that is the load-bearing point.

The stress side of the question has been measured too. A 2025 observational study tracked five hundred and fifty-seven horses across a six-and-a-half-hour event with multi-modal stress indicators. Stress behaviours rose alongside excessive whip use, and faecal cortisol approximately doubled by the end. The horses in that study are not dressage or jumping horses, but the physiological response to repeated whip use is not species-specific to a discipline.

The honest read for sport-horse riders is that the burden of proof on whip escalation sits with the practice, not with the data. Schooling tools that work through release of pressure rather than addition of pain have a stronger evidence base behind them.

What to do tomorrow

Saddle. Find a fitter who measures during movement, not only on a standing horse. Ask what the saddle is doing at trot, where the pressure peaks under increased girth tension, and how fit will be reassessed as your horse changes shape. Your own assessment of saddle position is wrong about forty-two percent of the time, regardless of experience.

Noseband. Check tightness with two fingers stacked flat on the dorsal midline, the bony bridge of the nose. Below roughly one and a half fingers, pressure rises sharply and stride length and airway function are measurably affected.

Bit. Bit width that fits your horse's mouth is mechanically supported. Wider-than-fit bits and high rein angles produce more lateral and posterior bit movement, with risk of contact at oral soft tissue. Deeper claims about specific mouthpiece materials and shapes outpace what the modern literature can establish.

Bridle type. Snaffle versus double is not the most determinative variable. Rider technique and noseband configuration carry comparable or larger weight.

Whip. The cleanest measurement to date shows the strike produces slowing, not the speed-up it is supposed to. The data does not generalise wholesale to every ridden context, but it shifts the burden of proof onto the practice. If you carry a stick on the assumption that escalation sharpens the response, the assumption has been tested and it failed.

Ridden behaviour. If your horse is showing conflict markers (mouth opening, head tossing, tail swishing, repeated head position changes), those are signals worth working up rather than schooling through. Gastric disease can drive the same markers, so a workup that includes both musculoskeletal and gastrointestinal evaluation is the supported path.

When to call a vet, a saddle fitter, or a trainer

Call a vet when ridden behaviour changes meaningfully over weeks rather than days, when your horse shows pain on palpation, when lameness or asymmetry is suspected, when RHpE-style markers persist after equipment changes, or when gastric disease is plausible. Musculoskeletal pain often co-occurs with poor saddle fit, so the workup needs both perspectives.

Call a saddle fitter at least annually, after meaningful changes in your horse's condition, after any structural change to the saddle, and as a first step when ridden behaviour changes.

Call a trainer for technique-mediated questions: contact quality, transitions, posture under load, lesson architecture. A 2025 paper found that instructor effects on rein tension are significant and consistent, which means contact is shaped by lesson design rather than by horse or rider in isolation.

The three roles overlap rather than substitute. Where a question lives in more than one of them, work through them in sequence.

Bottom line

Keep the noseband at two fingers, do not ride in deep flexion, treat the ridden horse pain ethogram as a screen rather than a diagnostic, measure saddle fit during movement, and remember that for treeless saddles, spurs, and training aids the verdict at three years is "we do not yet know" rather than "this works."

Anchor papers cited

1. Becard B, Sapone M, Martin P, Hanne-Poujade S, Babu A, et al. IMU quantification of saddle fit effects. Sensors (Basel). 2025. https://doi.org/10.3390/s25154712
2. Byström A, Egenvall A, Eisersiö M, Engell MT, Lykken S, Lundesjö Kvart S. Teaching approach effects on rein tension. Heliyon. 2025. https://doi.org/10.1016/j.heliyon.2025.e41947
3. Doherty O, Conway R, McGreevy P. Cadaver-head pressure under noseband tightness. Animals (Basel). 2025. https://doi.org/10.3390/ani15142141
4. Douglas M, Williams J, Randle H. Rider accuracy in assessing saddle position. Journal of Equine Veterinary Science. 2025. https://doi.org/10.1016/j.jevs.2025.105655
5. Dyson S, Pollard D. RHpE before and after diagnostic anesthesia. Animals (Basel). 2023. https://doi.org/10.3390/ani13121940
6. Dyson S, Pollard D. RHpE in 1,358 competition horses. Animals (Basel). 2024. https://doi.org/10.3390/ani14121831
7. Faithfull R, Lewis K, Drury E, McBride S. Double vs. snaffle bridle and conflict behavior. Animals (Basel). 2025. https://doi.org/10.3390/ani15121782
8. Fialová S, Kuřitková D, Sobotková E. FEI noseband measurement at competition. Animals (Basel). 2026. https://doi.org/10.3390/ani16030518
9. Galotti A, Eisersiö M, Yngvesson J, Lanatà A, Maglieri V, Palagi E, Baragli P. Rein tension and heart rate variability. Journal of Animal Science. 2025. https://doi.org/10.1093/jas/skaf146
10. Henshall C, McGreevy P, Shea G, Doherty O, Christensen JW, Fenner K, Warren-Smith A, McLean A. Commentary on lateral noseband measurement. Animals (Basel). 2026. https://doi.org/10.3390/ani16030412
11. Hopkins E, Whitrod S, Marlin D, Blake R. Tight nosebands and stride kinematics. Journal of Equine Veterinary Science. 2025. https://doi.org/10.1016/j.jevs.2025.105654
12. König von Borstel U, Kienapfel K, McLean A, Wilkins C, McGreevy P. Hyperflexion meta-analysis. Scientific Reports. 2024. https://doi.org/10.1038/s41598-024-72766-5
13. Loosley H, Rehbein LS, Lima EMM, Blake R. Roller padding under Pessoa training aid. Journal of Equine Veterinary Science. 2026. https://doi.org/10.1016/j.jevs.2026.105855
14. MacKechnie-Guire R, Clayton H, Williams J, Marlin D, Fisher M, Fisher D, Walker V, Murray RC. Snaffle vs. double bridle rein and headpiece forces. Animals (Basel). 2025. https://doi.org/10.3390/ani15071058
15. MacKechnie-Guire R, Murray R, Williams JM, Nixon J, Fisher M, Fisher D, Walker V, Clayton HM. Noseband type and tightness at trot. Equine Veterinary Journal. 2025. https://doi.org/10.1111/evj.14420
16. Marlin D, Randell O, Mayhew E, Blake R. Girth design and tension on saddle pressures. Animals (Basel). 2025. https://doi.org/10.3390/ani15172540
17. Maśko M, Sikorska U, Borowska M, Zdrojkowski Ł, Jasiński T, Domino M. Lunging aids and posture/temperature response. Animals (Basel). 2023. https://doi.org/10.3390/ani14010022
18. Murray R, Fisher M, Fairfax V, MacKechnie-Guire R. Saddle thigh block design. Animals (Basel). 2023. https://doi.org/10.3390/ani13132127
19. O'Connell E, Dyson S, McLean A, McGreevy P. Conflict behavior reframing. Animals (Basel). 2025. https://doi.org/10.3390/ani15030399
20. Pineau V, Ter Woort F, Julien F, Vernant M, Lambey S, Hébert C, Hanne-Poujade S, Westergren V, van Erck-Westergren E. Diet adaptation, gastric disease, and RHpE. Journal of Veterinary Internal Medicine. 2024. https://doi.org/10.1111/jvim.17223
21. Pollaris E, Hannes C, Haspeslagh M, Demey W, Teysen S, Boussauw B, Vlaminck L. Intra-oral bit movement (cadaver). Animals (Basel). 2025. https://doi.org/10.3390/ani15182648
22. Rodríguez-Segura BM, Chaves-Araya S, Estrada-McDermott JM, Saborío-Montero A, Wickens CL, Arias-Esquivel AM. Costa Rican horse parade welfare. Journal of Equine Veterinary Science. 2025. https://doi.org/10.1016/j.jevs.2025.105748
23. Sandberg A, Melani M, Flöhr A, Yngvesson J. Whip use in trotting racing. Frontiers in Animal Science. 2025. https://doi.org/10.3389/fanim.2024.1416503
24. Scholler D, Wittenberg J, Zablotski Y, May A. Tight nosebands and upper airway function. Veterinary Medicine and Science. 2024. https://doi.org/10.1002/vms3.1478
25. Smirnova KP, Frill MA, Warner SE, Cheney JA. Dynamic back shape during locomotion. Journal of the Royal Society Interface. 2024. https://doi.org/10.1098/rsif.2023.0644
26. Tilley P, Simões J, Sales Luis JP. 15-degree poll-flexion variation. Animals (Basel). 2023. https://doi.org/10.3390/ani13101714
27. Uldahl M, Mellor DJ. Regulatory integrity in horse sport. Animals (Basel). 2025. https://doi.org/10.3390/ani15131934
28. Wilkins C, Christensen JW, Doherty O, Fenner K, Freire R, Henshall C, McGreevy P. Comments on Clayton et al. 2024. Equine Veterinary Journal. 2025. https://doi.org/10.1111/evj.14548