Range of motion is the arc a joint travels between its two ends, and it is the only quantity underneath any statement about flexibility. Everything else in the category — tight, stiff, open, mobile, loose — is a description. This is the number.
It is usually written in degrees, sometimes in centimetres, and sometimes as a landmark you can name out loud: fingertips to the shin, knuckles to the floor, heel flat or heel lifted. All three are legitimate. Which one you should use depends less on precision than on whether you can take the same reading twice.
The short answer
- Range of motion (ROM) is how far a joint moves, expressed as a quantity.
- Passive ROM is the arc when something other than you does the moving — gravity, a strap, a hand, a clinician.
- Active ROM is the arc you can produce under your own muscular control.
- Active is bounded by passive. The gap between them is a reading in its own right, and often a more useful one than either number alone.
What range of motion measures
A joint's arc is set by several things at once, and they are not equally changeable.
Bone. The shape of the two surfaces, and how they are oriented, sets a hard limit. Hip socket depth and femoral neck angle vary enormously between people who are all perfectly healthy. Frank and colleagues, in Arthroscopy in 2015, pooled imaging studies of people with no hip symptoms at all and found cam morphology — a bump on the femoral head–neck junction that mechanically blocks deep hip flexion — in roughly a third of them. No stretching programme moves bone.
Capsule and ligament. The sleeve around the joint and the cords that reinforce it. These change slowly and mostly in response to sustained loading, not to a thirty-second stretch.
Muscle and its connective tissue. The part everyone means when they say flexibility. It is genuinely part of the answer, and it is a smaller part than the marketing implies.
Your nervous system. The point at which you stop is not a mechanical failure of tissue. It is a judgement, and a movable one. This is why you have more range warm than cold, more in the afternoon than at 7am, and more after two minutes of easy movement than after none.
So a single reading is a composite of an unchangeable structure and a changeable state, and it does not tell you which is which. That is not a flaw in the measurement. It is a fact about joints, and knowing it stops you spending a year stretching a femur.
Degrees, centimetres, landmarks
Three ways to write the number down, in descending order of laboratory precision and ascending order of how likely you are to actually do it.
Degrees, taken with a goniometer — two arms and a protractor lined up on bony landmarks. This is the clinical standard, and it is what the published normal tables are built from. It needs a second person for most joints and it needs the landmarks found the same way each time.
Centimetres, taken against a fixed reference. The weight-bearing lunge, where you measure how far your toe can be from a wall while your knee still touches it, is the best example: a wall does not move, a tape measure does not have an opinion, and the reading is a real length. The knee to wall test is that measurement, and it is one of only two readings in Limber's battery that produce a continuous number rather than a band.
Landmarks, named in plain words. Fingertips at mid-shin. Knuckles on the floor. Fingers overlapping behind the back. Less precise, and far more repeatable at home than a decimal place you invented while folded over.
The trade is not precision against sloppiness. It is precision against repeatability, and repeatability wins, because you are comparing your reading to your own reading six weeks from now, not to a laboratory.
Active range and passive range
Take the same joint twice. Lie on your back, loop a strap over one foot, keep the knee straight and pull the leg toward you until the stretch is firm but not sharp. That is passive. Now drop the strap and lift the same leg as high as you can with the muscles at the front of your hip doing all the work. That is active.
For most people the second number is well short of the first. That difference has a name in physiotherapy — the active–passive deficit — and it points somewhere useful. A large gap says your limit is control, not length, and more stretching adds arc you already cannot use. A small gap with a low number says length genuinely is the constraint.
Mobility vs flexibility works through what to do in each case. The point here is narrower: range of motion is a family of numbers, not one number, and saying which one you took is part of the reading.
The published normal tables
You will find tables of normal joint angles in every clinical textbook: shoulder flexion 0–180°, hip flexion 0–120°, ankle dorsiflexion 0–20°, and so on. They are useful. They are also weaker evidence than their confident presentation suggests, and it is worth knowing why.
The two most-cited primary sources are small. Boone and Azen published normal ranges in the Journal of Bone and Joint Surgery in 1979 from 109 male subjects. Roach and Miles, in Physical Therapy in 1991, did better — they used hip and knee data from the first National Health and Nutrition Examination Survey, a genuine population sample of adults aged 25 to 74 — and found the age-related decline was real but modest, on the order of a handful of degrees across five decades.
The tables reproduced in textbooks are largely consensus figures assembled by committees from sources like these, not percentiles from a large representative sample of your age and sex. Treat them as a sanity check on the order of magnitude and nothing more.
Grade: moderate for the population-sampled joints (hip, knee), thin elsewhere. A number outside the published band is a reason to look again, not a diagnosis.
How reliable a measurement is
This is the part almost nobody publishes and everybody needs.
Gajdosik and Bohannon reviewed goniometry in Physical Therapy in 1987 and set out the pattern that has held up since: the same tester measuring the same joint twice is reasonably consistent; two different testers measuring the same joint are noticeably less so; large joints measured in one plane do better than small joints or combined movements; and eyeballing an angle is worse than either.
The practical consequence is a rule you can apply this afternoon. Compare your readings only to your own readings, taken by you, the same way. A number from your physio's goniometer and a number from your bathroom floor are two different instruments, and the difference between them is mostly instrument.
It also sets a floor on what counts as a change. If your own repeat measurement wanders by a couple of centimetres, then a two-centimetre improvement is not evidence of anything. On the knee-to-wall scale Limber uses, one centimetre is worth about ten points near the middle of the range — which is nearly the whole of the twelve-point gap the app calls an imbalance. Small differences on a home test are frequently just the test.
What a number does not tell you
A range of motion reading is a measurement of an arc. It is not a measurement of health, of injury risk, or of how well you move.
It does not diagnose anything. Screening batteries built on movement quality have a poor record at predicting who gets hurt — Moran and colleagues' 2017 systematic review in the British Journal of Sports Medicine found Functional Movement Screen composite scores predict injury poorly, and Bahr's 2016 review in the same journal argued that this is a general property of screening tests rather than a fault of one battery.
And it says nothing reliable about pain. Stiffness and pain are different problems that arrive together often enough to be confused constantly. If you have pain that radiates into a limb, numbness, weakness, pain that followed a fall or an injury, or back pain with fever, unexplained weight loss or any change in bladder or bowel control, that needs a clinician who can examine you. Nothing on this page diagnoses or treats anything.
Taking your own readings
Five rules, and they are the whole method.
- Same protocol every time — same setup, same knee position, same cue, same landmark.
- Same conditions — always warm or always cold, and roughly the same hour of day.
- Both sides, recorded apart. An average of a left and a right hides the thing most worth knowing.
- Write the date on it. A reading without a date compares to nothing.
- Weeks between readings, not days. Range moves more slowly than the daily noise does.
Limber's battery is five of these readings scored on a 0–100 scale and rolled into one composite. You can run the same five for nothing on this site: the Range Score composes whatever you have measured and draws the ground you have not, dashed, rather than filling it in with a guess. It currently reaches four of the app's five score groups, because the spine group is fed by a test that is not built yet — on either surface.
More on the protocols in how to measure flexibility at home, and on what the resulting number means in what is a good flexibility score. The cluster this belongs to is mobility, measured.
Where the evidence is thin
Normal ROM tables. Widely printed, weakly sourced for most joints, and almost never broken down by the limb proportions and joint shapes that drive the number.
Home measurement reliability. Studied for a few tests — the weight-bearing lunge has published reliability, the sit-and-reach family has published validity — and essentially unstudied for the rest, which is why this site publishes which of its readings are proxies.
Range and outcomes. That more range makes you healthier, less injury-prone or better at anything is much less established than the category assumes. It has been argued in print that flexibility does not belong among the major components of physical fitness at all. We think measuring is still worth doing, because you cannot manage what you never wrote down — but we are not going to pretend that a higher number is self-evidently a better life.
Questions
What does ROM stand for?
Range of motion — the arc a joint travels between its two limits, written in degrees, centimetres, or against a physical landmark. It is the only quantity underneath words like tight, stiff or flexible, all of which are descriptions rather than measurements.
What is the difference between active and passive range of motion?
Passive range is how far the joint goes when something other than you moves it: gravity, a strap, a hand. Active range is how far you can move it yourself, with the muscles on both sides of the joint doing their jobs. Active is always less than or equal to passive, and a large gap between them means your limit is strength and control rather than tissue length.
How is range of motion measured at home?
With a repeatable landmark rather than a protractor. A wall, a floor and a fixed starting position give you a reading you can take identically in six weeks, which is the only comparison that matters. Measured lengths in centimetres, like the knee-to-wall test, are the most precise home readings available; banded landmarks are less precise and far more repeatable than a guessed angle.
What is a normal range of motion?
The published tables give figures like 0–120° of hip flexion or 0–20° of ankle dorsiflexion, but they come from small samples or from committee consensus, and they do not account for the joint shape and limb proportions that legitimately vary between healthy people. A reading outside the band is a reason to look more carefully, not a finding.
Does a low reading mean something is wrong?
Not on its own. Bone shape, limb proportions, age, warmth, time of day and habit all move the number, and imaging studies show that structural variants which limit range are common in people with no symptoms whatsoever. If you have pain, numbness, weakness, or symptoms that followed an injury, see a clinician who can examine you rather than reading a test result off the internet.
How much does a reading have to change to be real?
More than the test's own noise. The same person measuring themselves twice will not get exactly the same figure, and on a centimetre-scale test that wobble can be worth ten points on a 0–100 scale. Take readings weeks apart under the same conditions, and treat a change smaller than your own repeat variation as no change at all.