Squat depth mobility sounds like one quality and is three joints asked for a lot of range at the same instant. The squat itself cannot tell you which of the three ran out; it only tells you that something did. So the useful work is not more squatting — it is finding out which joint is the ceiling, because two of the three respond to training and one of them is the shape of your femur.
This page gives the measured demands, the controlled demonstration that the ankle governs the shape above it, the reason hip anatomy sets a hard limit for some people, and what Limber's own readings can and cannot say about any of it.
The short answer
- A full squat has been measured at roughly 95 degrees of hip flexion, 155 of knee flexion, 38 of ankle dorsiflexion.
- Restrict the ankle in a lab and the squat changes immediately. The ankle governs the shape.
- Stretching the calf buys a few degrees, of questionable clinical importance.
- Cam-type hip morphology turns up in about a third of people with no symptoms. Bone does not stretch.
- The deep squat test is a four-band screen and cannot name the joint that stopped you.
- Before more stretching: stance width, foot angle, a heel wedge.
What a full squat costs
Hemmerich and colleagues recorded hip, knee and ankle kinematics during high-range daily activities in the Journal of Orthopaedic Research in 2006, in a population where deep squatting is ordinary rather than a gym skill. A full squat took roughly 95 degrees of hip flexion, 155 degrees of knee flexion and about 38 degrees of ankle dorsiflexion.
Grade: solid for the measurement. These are recorded joint angles, not estimates, though they come from one study in one population and the numbers move with foot position and squat style.
Thirty-eight degrees is the striking one. A running stride uses perhaps 10 to 15. Walking uses less. Most adults who sit for a living have never asked their ankles for anything close, and the ankle is the joint at the bottom of the chain, which is why it decides so much.
Squat depth mobility, joint by joint
The ankle. Trainable, measurable in real units, and the one with a controlled demonstration behind it. Measure it directly rather than inferring it.
The knee. Almost never the limit in a healthy adult. A knee that will not fold is a different problem from a stiff squat and belongs with a clinician who can examine it.
The hip. Partly trainable, partly not. Soft tissue responds; the shape of the socket and the head of the femur does not.
The spine. A squat that collapses forward at the bottom is often reporting an ankle, not a back. Fixing the position at the wrong end of the chain is the commonest wasted month in this subject.
There is one more reason to start at the ankle rather than the hip. Ankle range is trainable, cheaply testable and expressed in real units, so you can tell within two minutes whether it is your ceiling. Hip range is partly bone, partly soft tissue, and no home test separates the two. What a home screen can do is read the two hip positions it can score and compare your sides on the one it cannot: the hip mobility test is a 30-second squat hold, a hip flexor check and a seated rotation comparison, and the hips number it returns comes from the first two only. Working on a joint you can measure, in units you can compare, is the difference between six weeks of progress and six weeks of guessing.
Test the ankle first because it is the cheapest thing to rule in or out. The knee-to-wall test reads it in centimetres, one side at a time: 2 cm reads 15 on Limber's scale, 4 cm reads 35, 6 cm reads 55, 8 cm reads 70, 10 cm reads 85, 12 cm reads 95, interpolated between and flat outside.
The ankle, demonstrated
Macrum and colleagues, in the Journal of Sport Rehabilitation in 2012, artificially restricted ankle dorsiflexion in healthy people and watched the squat change immediately: less knee travel, more forward trunk lean, altered muscle activation. Nothing about the hips or the spine was touched. The shape above changed because the joint below was shortened.
Grade: solid for the mechanics. A controlled manipulation with an immediate, measured effect.
Read it precisely, because it is easy to over-read. It shows that restricting an ankle changes a squat. It does not show that raising a restricted ankle by a few degrees changes anybody's squat, or their numbers, or their injury rate. That trial has not been run.
And the intervention is modest. Radford and colleagues reviewed the trials of static calf stretching in the British Journal of Sports Medicine in 2006: pooled, it produced a statistically significant gain in dorsiflexion of a few degrees, which the review itself called of questionable clinical importance. Grade for the intervention: thin.
Which is why the cheap changes come first. A wider stance, feet turned out a little, and a small heel raise all reduce the dorsiflexion the squat demands, immediately and for free. Ankle dorsiflexion explained covers the joint in full.
Bone is not a tissue
Some squats stop for a reason no routine reaches.
Frank and colleagues reviewed imaging of people without symptoms in Arthroscopy in 2015 and found cam-type morphology — extra bone at the head-neck junction of the femur, which mechanically blocks deep flexion and inward rotation — in roughly a third of asymptomatic volunteers and over half of asymptomatic athletes. Where bone meets bone, the joint stops, and the femur does not lengthen with practice.
Grade: reasonably solid for the prevalence, unsettled for what it means. A scan finding in someone with no complaint is not a diagnosis, which is exactly why Griffin and colleagues' Warwick Agreement, in the British Journal of Sports Medicine in 2016, defines hip impingement syndrome as symptoms, clinical signs and imaging together, never imaging alone.
The practical version: if your squat feels blocked at the front of the hip in a hard, bony way rather than stretched at the back in a soft one, change the stance before you change the routine. Deep hip or groin pain that catches, gives way, persists or wakes you belongs with a clinician who can examine you, not with more depth work.
The reading, and its error bar
Two numbers, and both come with limits worth knowing before you read anything into them.
The deep squat test is four bands, captured as a thirty-second hold: I can't reach parallel reads 15, parallel, heels lifted reads 40, full squat, heels down reads 70, a comfortable resting squat reads 90. Note what separates the top two — comfort, not depth. Above band 3 the test has stopped measuring how low you go and started measuring whether the bottom is somewhere you can stay.
The bands are 20 to 30 points apart, so the smallest change this test can express is a whole band. Real progress inside a band shows up as nothing.
The knee-to-wall is the opposite problem: too fine rather than too coarse. One centimetre is worth 5 to 10 points, and Powden, Hoch and Hoch reported in Manual Therapy in 2015 that the test's minimal detectable change sits around one to two centimetres. Bennell and colleagues had established its reliability in the Australian Journal of Physiotherapy in 1998, good enough to compare two of your own readings. So a small difference between attempts is the instrument, not you. Same wall, same foot, same shoes or none.
Neither reading predicts anything. Moran and colleagues found in 2017 that movement-screen composites predict injury poorly, and Bahr argued in 2016 that this is a property of screening in general.
What the score does with a squat
One squat reading is the busiest number in Limber's battery. It feeds three areas at once.
Hips are the deep squat at 50 per cent plus the hip flexor test at 50 per cent. Glutes are the deep squat at 100 per cent, and the app states that it is a proxy rather than a glute measurement. Lower back is the toe touch at 50 per cent plus the deep squat at 50 per cent.
So a band change on this one test moves three areas and two of the five score groups. That is worth knowing before you read a jump in your Range Score as five separate improvements. A worked survey, if you want to see the arithmetic: this one reads 58 over four of five stations, with the spine left blank because the test that would feed it is not built.
The engine also acts on the ankle gap. Limber calls an imbalance at 12 points, and on the knee-to-wall that is about 1.2 cm through the middle of the scale. When it fires, the plan adds one extra move for the low side until the gap falls under 6.
Mobility for lifters puts the same joints under a bar, rest day mobility covers the day the depth work belongs on, and deep squat mobility is the anatomy. The route from where you are to a resting squat is deep squat progression, and the cluster sits under mobility for training.
Questions
Why can I not squat deep with my heels down?
Most often the ankle, sometimes the hip, and the squat itself cannot tell you which. Measure the ankle with the knee-to-wall before you spend a month on anything: a full squat has been measured at about 38 degrees of dorsiflexion, far more than walking or running ever asks. If the ankle reads well and the block feels hard and bony at the front of the hip, stance and foot angle will do more than stretching.
Will stretching my calves let me squat deeper?
A little, at best. Pooled trials of static calf stretching found gains of a few degrees of ankle dorsiflexion, which the review itself called of questionable clinical importance, and no trial has shown that raising a restricted ankle changes anyone's squat. A wider stance, feet turned out and a small heel raise reduce the demand immediately and cost nothing.
Is my hip anatomy stopping me?
Possibly, and there is no way to know from a squat alone. Cam-type hip morphology turns up in roughly a third of people with no symptoms at all, and where bone meets bone the joint stops regardless of how much time you spend on a mat. A scan finding in someone with no complaint is not a diagnosis, which is why the clinical definition requires symptoms, signs and imaging together.
How deep should a squat be?
However deep you can hold with the position intact, which is a different answer for different skeletons. Limber's top band is not the lowest squat but the most comfortable one: band 3 is a full squat with heels down, and band 4 is a squat you can rest in for thirty seconds. Depth beyond a position you can own is a demonstration rather than a capacity.
Why has my squat score not moved?
Probably because the test cannot see the change. Four bands 20 to 30 points apart mean real progress inside a band reads as nothing, and going from the bottom of band 3 to the top of it is invisible. The knee-to-wall has the opposite failing and will show you movement the squat band conceals.
Does the app have a squat programme?
Not a sport-specific one. Ten programme tracks are drawn, including Hips Unlocked and Ankles and Feet, and there is no squat track, no lifter track and no sport packs. What the engine does is build a routine from the areas you choose, your readings, your equipment and your flagged movements, with a stated reason for every move it picks.