A middle splits progression runs into a wall the front splits does not, and it is worth saying at the top: of the two splits, this is the one where individual hip anatomy most often decides the outcome. The position drives the thigh bones straight out sideways into the rim of the socket, and how much room there is before they meet varies enormously between people.
That is not a reason to skip it. It is a reason to hold the goal loosely and measure rather than assume.
The short answer
- The demand is adductor length plus hip abduction with the right rotation.
- Socket orientation varies between people and is not trainable. This is where the middle splits ends for some.
- Turning the feet and knees up rather than forward changes the position substantially.
- No trial exists on any splits programme. Everything here is mechanism plus borrowed dose research.
- We print no date, and our own ten-week middle-splits track is not built. Stated below.
A middle splits progression
Four stations, and the sideways rule beneath them.
Station one — the components.
- Adductors, long: a seated straddle with a long spine, hands forward on the floor, hinging from the hips rather than rounding. 45 seconds.
- Adductors, short: a Cossack shift — feet wide, sink to one side with that knee bent and the other leg straight, toes up. This loads the adductors near their end range with something to do, which passive stretching does not. Six shifts each side.
- Hip abduction with rotation: stand tall and take one leg out to the side as far as it will go under its own power, slowly. Six per side. What you can produce actively is the part you own.
Station two — the wide squat. Feet wide, toes turned out, sink between them. Elbows inside the knees, gently opening. 45 seconds. This is the most useful single position in the progression because it loads the abduction under body weight without asking for the full split shape.
Station three — the supported straddle. Against a wall, legs up and apart, letting gravity do the work over a couple of minutes. Or the same thing prone, on the forearms, sliding out.
Station four — depth, with support under the hands. Blocks or a chair, weight shared, sliding a little further as the position allows. Never as a cold test of willpower.
The rule that governs all of it: as the legs go wider, the knees and feet should point up toward the ceiling, not forward at the floor. A middle split with the feet forward is a different joint position, it is much harder on the knees, and it runs the thigh into the front rim of the socket sooner. Rotating the whole leg so the kneecap faces up is the single change that most often adds depth.
Why anatomy dominates here
The front splits mostly asks for flexion and extension, which the hip joint is shaped to do. The middle splits asks for abduction, and abduction is where the top of the thigh bone meets the rim of the socket soonest.
Frank and colleagues reviewed imaging of people without symptoms in Arthroscopy in 2015 and found cam-type hip morphology — extra bone at the femoral head-neck junction — in roughly a third of asymptomatic volunteers and over half of asymptomatic athletes. Acetabular depth and orientation also vary widely between individuals, and both determine how far the leg travels before bone meets bone.
Grade: reasonably solid for the anatomical variation, unknown for how many people it limits. Nobody has quantified what share of adults cannot reach a middle split for bony reasons, and anybody telling you the answer is anecdote is at least being honest about the source.
The signal is again the sensation. A stretch is felt as length along the inside of the thigh. Bone contact is a hard pinch at the front of the hip or deep in the groin, it does not soften with time, and it is a stop rather than a target.
Groin pain that catches or clicks painfully, pain persisting after the session, pain radiating down the leg, numbness or weakness is a matter for a clinician who can examine you. Groin injuries in particular are a recognised area where a careful examination beats a guess — Weir and colleagues published an agreed terminology for groin pain in athletes in the British Journal of Sports Medicine in 2015 precisely because the region is so often mislabelled. Nothing here diagnoses anything and nothing here is treatment.
The dose, borrowed and labelled
There is no adductor dose literature worth the name and no splits trial at all. What follows is transferred from hamstring research and should be treated as an approximation.
Bandy and Irion, in Physical Therapy in 1994, found 30 seconds beat 15 and matched 60 over six weeks in 57 people; their 1997 follow-up, with 93 people, found one 30-second hold a day matched three. Thomas and colleagues, in the International Journal of Sports Medicine in 2018, found total accumulated weekly time mattered more than how it was divided.
Grade: reasonably solid for hamstrings, and this is an extrapolation. Different muscle group, different joint action, different training duration, and no study of anybody pursuing a large-range position over months.
Two things are worth adding on top of the borrowed dose, both with better rationale than pure passive stretching:
Load the range. Cossack shifts and wide squats put the adductors under load near their end range. Afonso and colleagues' review in Healthcare in 2021 found strength training and stretching produced comparable range improvements, on a small and varied set of trials that the authors themselves flagged as a limitation. Grade: mixed — but for a position you have to be able to hold under your own weight, loading has an argument that passive hanging does not.
Contract-relax helps acutely. Hindle and colleagues reviewed the technique family in the Journal of Human Kinetics in 2012 and found larger acute gains than passive stretching. Its textbook explanation is another matter: Sharman, Cresswell and Riek argued in Sports Medicine in 2006 that the reflex inhibition it is supposed to exploit does not appear in the electromyography. The technique works; the story about why does not.
What we will not tell you
A date. No trial has measured how long a middle split takes for anyone, the between-person variation is large, and the ceiling is anatomical for some. A countdown would be invented precision. A range is a different thing, and the splits timeline prints one — six-week training blocks and a wide spread of months, from your current gap to the floor and how often you train — with the bone limit named on the page, since hip shape ends this split for some people whatever they do.
And the admission that goes with it. Limber's engine has a splits goal. Its programme library lists a ten-week middle-splits track that is not drawn, alongside an eight-week front-splits track in the same state. The app names them as not built and gives them no tap target rather than shipping a repeated week as a progression.
The reason is countable. The move catalogue holds 56 stretches, and the app's ten body areas do not include the adductors at all — the closest thing in the catalogue is a single Cossack shift filed under hips. Ten honest weeks of graded adductor work cannot be built from that. When the catalogue is deep enough the track will exist; until then, claiming it does would be the countdown by another route.
That it is a health goal. It is not. It is a skill, and a good one to want. Nuzzo argued in Sports Medicine in 2020 that flexibility has weak justification as a major component of physical fitness at all, and that argument is worth reading even by people who intend to keep stretching. Grade: a serious minority position, well argued.
Measure what you can
There is no middle-splits reading in any battery, ours included, and inventing a score for it would be exactly the thing this page is complaining about. What can be measured are the neighbouring readings that move alongside it.
The hip flexor test reads hip extension on each side in three bands — 25 for a thigh above the bench line, 60 for level, 90 for hanging below. It is the coarsest instrument in the battery and the app says so. Two limits worth repeating: it is not a valid measure of hip extension unless pelvic tilt is controlled, per Vigotsky and colleagues in PeerJ in 2016, so press the small of your back flat; and three bands means the smallest gap it can show between sides is 30 points, while 12 is called an imbalance, so any difference at all is flagged and should be re-measured rather than acted on.
The deep squat test reads the wide-stance position that the middle splits is built on top of, in four bands over a thirty-second hold: 15 for not reaching parallel, 40 for parallel with the heels up, 70 for a full squat heels down, 90 for a squat you can rest in.
Take them three weeks apart under matched conditions. Front splits progression covers the other direction, the pancake stretch is the seated relative of this position, and the cluster is goals.
Questions
How long does it take to get the middle splits?
Nobody knows. There is no trial of any splits programme, so every published timeline is extrapolation or anecdote. Hip socket shape also sets a real ceiling for some people, which means the honest answer varies more between individuals here than for almost any other flexibility goal.
Why do my hips pinch in the middle splits?
Usually because the thigh bone has met the rim of the socket. That is felt as a hard pinch at the front of the hip or deep in the groin rather than as a stretch along the inside of the thigh, and it does not soften with time in the position. Rotating the legs so the kneecaps face up often changes it; pushing harder does not.
Should my feet point up or forward in the middle splits?
Up. As the legs widen, turning the whole leg so the kneecap and foot face the ceiling changes where the thigh bone sits in the socket and is easier on the knee. Feet forward is a different joint position that meets bone sooner and asks more of the knee than of the hip.
Are the middle splits harder than the front splits?
For most people, yes, because abduction runs into bone sooner than flexion and extension do. It is also the position where individual anatomy most often decides the outcome, which is why coaching experience varies so widely on how long it takes.
What is the best stretch for the middle splits?
A wide squat with the toes turned out is the most useful single position, because it loads the abduction under body weight rather than only hanging in it. Cossack shifts add the loaded end-range work. Passive straddle holds are worth doing and they are not the whole plan.
Does Limber have a middle splits programme?
Not today, and the app says so. The ten-week middle-splits track is listed as not drawn, because the 56-move catalogue has no adductor area at all and one Cossack shift filed under hips. Publishing a repeated week and calling it a progression is what we are refusing.