Neck stretches are reasonable, low risk, and not a correction of anything. That is the whole honest claim, and almost every page on this subject makes a much larger one.
The neck is where a screen day lands and where the posture business has done its best trade. What follows is the seven movements in Limber's catalogue, their real holds, what to feel, when to stop, and a plain statement of what the evidence does and does not support.
The short answer
- Gentle range work for the neck is low risk and worth doing for how it feels.
- It is not a correction. Nothing here changes the shape of your neck or the position of your head.
- The catalogue holds seven neck movements. All are under a minute and most need nothing.
- Full neck circles are not among them, deliberately. Dizziness, visual change or nausea in any neck position is a reason to stop and be examined.
- The strongest evidence in this region is for two minutes a day of resistance training, not for stretching.
The neck stretches, one by one
Seven movements, with the app's own defaults. Nothing here needs kit except the wall nod, which needs a wall.
| Move | Type and hold | Kit |
|---|---|---|
| M010 Neck Side Tilt | hold, 30 s per side | none |
| M045 Chin Tucks | dynamic, 40 s | none |
| M046 Levator Scapulae Stretch | hold, 30 s per side | none |
| M047 Anchored Upper-Trap Stretch | hold, 30 s per side | none |
| M048 Scalene Stretch | hold, 30 s per side | none |
| M049 Neck Half-Circles | dynamic, 40 s | none |
| M050 Suboccipital Wall Nod | hold, 40 s | wall |
M010 Neck Side Tilt. Sit tall, let the ear drift toward the shoulder, and let the arm on the stretched side hang heavy. Thirty seconds each way. The commonest error is lifting the opposite shoulder to meet the ear, which removes most of the movement.
M045 Chin Tucks. Draw the chin straight back over the throat — back, not down. Down is a different movement and it is the usual mistake. This one is cleared for warm-ups and works as a dynamic movement rather than a hold: one repetition per breath cycle, forty seconds' worth. Ease off at any pinching at the base of the skull.
M046 Levator Scapulae Stretch. Turn the head about forty-five degrees away from the side being stretched, then look down toward the armpit. The levator runs from the upper neck to the top corner of the shoulder blade, and the combination of turn and tilt is what reaches it. Thirty seconds each side.
M047 Anchored Upper-Trap Stretch. The side tilt again, with the hand on the stretched side anchored under the seat of your chair or held behind the back. Anchoring the shoulder down is the whole point — without it the shoulder rises and you feel very little. Thirty seconds each side.
M048 Scalene Stretch. Tilt the head away and then turn the face slightly upward on the stretched side. The scalenes run down the side and front of the neck to the first ribs. This one wants less force than the others and more patience.
M049 Neck Half-Circles. From one shoulder, chin sweeping down across the chest, to the other shoulder, and back. Down only. Forty seconds. The upper half of the circle is deliberately absent and the next section explains why.
M050 Suboccipital Wall Nod. Stand with the back of the head against a wall, then nod as if agreeing very slightly, keeping the head in contact. It works the small muscles at the base of the skull rather than the long ones. Forty seconds. Ease off at any sharp pain or dizziness.
Dose, and what to feel
Holds run 20 to 60 seconds. Dynamic movements are one repetition per breath cycle, six to twelve repetitions. Breathe in for four seconds and out for six, which paces the repetitions for you and stops the whole thing turning into a rush.
The target sensation is a mild stretch — a broad pull along a line, on the side you meant to work. Not sharp. Not a pinch. Not anything that produces a symptom in the arm or the hand.
None of the seven is set up for PNF, the contract-and-relax method used on some of the leg and chest positions. That is deliberate. Organising a forty per cent contraction against a stretched neck from a web page is not a thing this site is going to do.
Frequency matters more than length here, and that argument is developed in text neck stretches, which is the phone-specific version of this page.
Circles, dizziness and stopping
This is the one safety point that genuinely belongs on a neck page, and it is why the catalogue has half-circles rather than circles.
A full neck circle takes the joint into combined extension and rotation at the end of its range, repeatedly, with the head's own weight behind it. Nothing about the lower half of the circle is a concern. The upper half is where the position stacks extension on rotation, and it is the position most likely to produce symptoms in people who have any reason to be sensitive to it. The trade is simple: the top half of the circle adds nothing you cannot get from a chin tuck and a rotation done separately, so it is not in the catalogue.
More important than the movement itself is the rule about what you feel.
Dizziness, light-headedness, visual disturbance, nausea, ringing in the ears, numbness around the mouth or unsteadiness, produced or worsened by a neck position, is a reason to come out of the position and get examined. Not to push through, not to try the other side, not to repeat it to check. This is uncommon and it is not a reason to be afraid of moving your neck. It is a reason to know the list.
The general referral line applies too. Pain that persists, wakes you at night, radiates into a limb, comes with numbness or weakness, or followed a fall or an injury is a matter for a clinician who can examine you. Nothing here diagnoses anything and nothing here is treatment.
What the evidence supports
Stated plainly, because the honest summary is short.
The neck literature is low quality almost everywhere you look. The Cochrane reviews of exercise for mechanical neck disorders found some benefit for specific strengthening work, and graded the evidence behind it as low quality. That is the ceiling of the field, not a stray weak result inside it.
Blanpied and colleagues' neck pain clinical practice guideline in the Journal of Orthopaedic & Sports Physical Therapy in 2017 recommends exercise — including both range work and strengthening — across most presentations of mechanical neck pain, and is explicit about how much of the underlying research is modest.
The posture explanation has repeatedly failed its own test. Richards and colleagues, in Physical Therapy in 2016, measured sitting neck posture in more than 1,000 adolescents, sorted them into clusters statistically rather than by eye, and found no association with neck pain. Grob, Frauenfelder and Mannion, in the European Spine Journal in 2007, asked whether the curvature of the cervical spine differs between people with and without neck pain and found no association either.
Grade: weak for stretching as a treatment. Reasonably solid for the negative on posture. Moderate at best for exercise in general, on low-quality evidence.
So what is left for a stretch? A change of position, a movement into a range a screen day never visits, and the fact that it feels good afterwards. That is a modest claim and it is a true one, which is more than the alternative.
The stronger finding is not a stretch
Printed at full strength, on a page about stretching, because the reader deserves it before the caveats.
Andersen and colleagues, in Pain in 2011, gave office workers with frequent neck and shoulder pain two minutes a day of progressive resistance training and measured a real reduction over ten weeks. A randomised trial, a clinical outcome, in this exact population, at a dose smaller than the time you would spend on the seven movements above.
Nothing in the stretching literature for this region comes close to it. If your neck aches at work and you can only do one thing, the evidence says make it that one, and treat the range work as the pleasant addition rather than the main event. The same argument is made from the desk side in tech neck explained.
The neck has no test of its own
The admission, since this page has spent its length grading other people's.
Limber does not measure your neck. The battery is five built tests and none of them is a neck test. The neck area in the score is fed by the shoulder reach at a weight of 1.0 — the whole of it, from a test of the arms — and the app names that as a proxy rather than presenting it as a reading. That is a compromise, published, not a measurement.
The segment underneath, the thoracic spine, is worse off: its test is designed and not built, so the Spine group never scores at all. Four of five stations is the honest ceiling of the whole battery. Thoracic spine mobility sets out that gap in detail, and upper back tightness covers the complaint that sits between the two.
What you can take today is the shoulder mobility test — the Apley scratch, both sides, five bands, reading the chain the neck sits on top of — and the Range Score, which composes whatever you have measured and draws the rest as unmeasured ground. The rest of the region is under shoulders and upper back.
Questions
Do neck stretches actually help?
They help the way a change of position helps: a movement into a range a held posture never visits, with an immediate and short-lived effect on how the neck feels. What they do not do is correct anything, and the trial evidence for stretching as a treatment for neck pain is weak. The best-supported intervention in office workers is a small daily dose of resistance training instead.
How long should I hold a neck stretch?
The app's own defaults are 30 seconds for the side tilt, the levator, the upper trap and the scalene positions, and 40 seconds for the wall nod. Holds anywhere from 20 to 60 seconds are within the range the general stretching research supports, and a 30-second hold has matched 60 in trials on other muscle groups. Breathe in for four seconds and out for six rather than counting.
Are neck circles safe?
Full circles are not in this catalogue, and half-circles are. The lower half of the circle is unremarkable; the upper half stacks end-range extension on rotation and adds nothing you cannot get more precisely from a chin tuck and a separate rotation. If any neck position produces dizziness, visual change, nausea or unsteadiness, come out of it and see a clinician who can examine you.
Which stretch is best for the muscle between neck and shoulder?
That is usually the upper trapezius or the levator scapulae, and the catalogue has one for each. The anchored upper-trap stretch is a side tilt with the hand on the stretched side held down, 30 seconds. The levator position adds a turn of about 45 degrees away from the stretched side and then a look down toward the armpit, also 30 seconds.
How often should I do them?
Often and briefly beats rarely and long for a complaint driven by time in one position. Two or three short passes through a working day is a sensible default, and the evidence for that pattern is mechanistic rather than trialled — it follows from what a held position does, not from a study of this schedule. Total weekly time is the variable the stretching literature actually identifies.
Can Limber score my neck?
Not directly. There is no neck test in the battery, and the neck area of the score is inferred entirely from the shoulder reach test, which the app labels as a proxy. The thoracic spine below it has no reading either, because its test is designed and not built, so the honest ceiling of the whole battery is four of five stations.