How long does rotator cuff tendonitis take to heal? What the evidence supports

How long does rotator cuff tendonitis take to heal?

Months rather than weeks, and no honest page can narrow that to your own number. That is not us dodging the question. The prediction has been attempted and it did not work: a systematic review of five low-quality prognostic studies in 387 adults receiving physiotherapy found a lack of any valid and useful model for predicting outcomes in adults with pain from rotator cuff tendinopathy (Braun and colleagues, 2016, as reported by Desmeules and colleagues, 2025). Somebody looked, carefully, and came back without a model.

What you are left with is three published figures that get quoted as though they were one. They are not, and the mix-up costs people their programs. Read six to twelve weeks on a clinic blog, still be sore at week ten, and the obvious conclusion is that the exercises have failed. Usually it is the wrong conclusion.

This page keeps the three apart. It also says where the two useful dates come from and what usually improves first, so you have something to measure. And it says when waiting is the wrong call. The program the clock is running on is the rotator cuff tendinopathy program, and if you are trying to get back under a barbell inside that time, the gym with a rotator cuff injury is the page for the training side.

One number comes before all of them, and it falls far earlier than anything below. The NHS advises seeing a doctor or physiotherapist if shoulder pain is getting worse, if it has not improved after 2 weeks, or if it is very difficult to move your arm or shoulder. A long horizon describes the ordinary course of a shoulder that is slowly improving. It is not a reason to sit out one that is not improving, and nothing further down this page asks you to wait for a date before getting help.

How long does a rotator cuff injury take to heal?

This is the phrase most people type, and it has to be taken apart first, because rotator cuff injury covers three situations and the clocks are nothing like each other. Picking the wrong one is how somebody with a torn tendon ends up waiting out a tendinopathy timeline.

The first is the subject of this page: a cuff tendon that has become painful and sensitive to load over weeks or months, with no injury behind it. That is what the 2025 clinical practice guideline covers. Its own scope takes in rotator cuff tendinopathy with or without calcification along with partial-thickness tears, and excludes other rotator cuff diagnoses such as full-thickness tears (Desmeules and colleagues, 2025).

The second is a full-thickness tear, where the tendon is through. That has its own evidence, its own decision about surgery and its own timings, and they are on rotator cuff tear exercises and the decision about surgery. Nothing on this page is a timeline for it.

The third is a shoulder that has been repaired. After shoulder surgery, including a rotator cuff repair, or after a shoulder fracture or dislocation, follow your surgeon's or physio's plan, and do only the movements and range they have cleared. The staging and the published figures for that belong to rehab after rotator cuff repair, and this page gives none of them on purpose: a recovery figure borrowed from the wrong operation is worse than no figure at all. One line in the warning list at the foot of this page is for you specifically, the one about infection since your operation, and it does not wait for a timeline.

A fourth situation fits none of those three. If your shoulder pain started with a fall, a hard pull or another injury, and the arm has been weak since, that is a same-day matter rather than a waiting game, and the warning signs at the foot of this page say so in full. A cuff tear caused by an injury is referred urgently rather than managed with time, and the UK shoulder surgeons' pathway asks for the next available clinic appointment (Kulkarni and colleagues, 2015).

How long ago it happened does not soften that. If the fall was six weeks ago and the arm has been weak ever since, the dates further down this page are not the ones to measure it against, and nothing about the delay makes an assessment less worth having. The 2025 guideline puts it in its own reasoning rather than in a numbered recommendation: imaging should be used where there has been trauma, or where a significant structural lesion such as a full-thickness tear is suspected, as well as after nonsurgical treatment has failed (Desmeules and colleagues, 2025). Surgeons also weigh how long it has been when they decide what to offer, and the published series disagree with one another about how much a delay costs (Gutman and colleagues, 2021; Bjornsson and colleagues, 2011). When the evidence argues with itself like that, it is a reason to be looked at early, not a reason to work a deadline out from a page.

Rotator cuff tendinitis recovery time: three figures that get mixed up

So here they are one at a time, with what each one is a measurement of. Tendinitis and tendonitis are the same word with two spellings, and both mean the same shoulder.

The first is a program length. A review of 14 trials of exercise for rotator cuff tendinopathy concluded that most programs should show a clinically significant result by 12 weeks (Littlewood and colleagues, 2015). Read that as written. It is the point by which a program that is going to work should have shown something, which is why it is also when to judge one. It does not say you will be finished at 12 weeks, and it is not a prediction about any individual. The same number turns up second-hand in a 2021 review of rotator cuff rehabilitation, where it has become a minimum duration of 12 weeks for an individualized, progressive program, and that review is passing on this one (Boland and colleagues, 2021). The minimum wording is Boland's restatement, not what the 2015 review says.

The second is a review trigger, and it happens to be the same number. The 2025 guideline uses a maximum of 12 weeks of appropriate nonsurgical care as the point at which two different things become reasonable: imaging, if symptoms have not resolved or improved by then, and a referral to a musculoskeletal physician specialist for someone with severe and persistent pain or disability (Desmeules and colleagues, 2025). That is a date for a decision, not a date for being better. Both halves of it are written for a shoulder with no injury behind it and no tear suspected; where there has been an injury or a tear is suspected, the same guideline puts the scan earlier, as the section above says.

The third is the recovery horizon, and it is not 12 weeks at all. The NHS puts it at 6 months or longer to recover from shoulder pain, and advises trying shoulder exercises for 6 to 8 weeks to stop the pain returning. OrthoInfo, the patient site of the American Academy of Orthopaedic Surgeons, describes gradual improvement and a return to function with nonsurgical treatment, while saying complete recovery may take several weeks to months. Healthify New Zealand says most shoulder pain settles over a few weeks, which is a wider group of shoulder problems than this one. Australia's healthdirect, on rotator cuff injury, gives no recovery timeframe at all for treatment without surgery, so nobody can quote it for one; what it does say is to see a doctor if shoulder and arm pain has not gone after a few weeks.

Put the three side by side and most of the advice online falls apart. A number offered to you as your recovery time is almost always one of the first two dressed up as the third.

Why can nobody predict your rotator cuff tendonitis recovery time?

Because somebody has already tried and the work did not produce a usable answer. That is not the same as nobody knows, and it is more useful.

What the 2025 guideline reports is the review described at the top of this page: five prognostic studies, 387 adults having physiotherapy, and no valid, useful model for predicting how an individual will do (Braun and colleagues, 2016, as reported by Desmeules and colleagues, 2025). Five studies is a thin base, and they were low-quality ones, which is why the result is an absence of a model rather than a model that performs badly.

The same guideline does report directions, and a direction is worth having even without a number. It describes moderate evidence that being over 55, and perceiving your job demands as high, are associated with a higher risk of persistent pain.

Those directions appear here without the figures that came with them, on purpose. The review they come from is about the chronification of non-traumatic shoulder pain in general, the guideline reports that these two factors rest on two trials inside it, and we could not read the review ourselves. A risk figure resting on two trials nobody here has read is exactly the kind of thing that gets quoted back at you as a prognosis. The same review is also the source of a short list of apparently protective factors, one of which is counterintuitive enough that we are not repeating it until somebody here has read the trials underneath it.

Two trials give a sense of the timescale the researchers themselves expected. GRASP randomized 708 adults with rotator cuff disorders across 20 UK NHS trusts and followed them for 12 months. It found no difference between progressive exercise and best practice advice, and none between a corticosteroid injection and no injection (Hopewell and colleagues, 2021). That is not a finding that nothing is worth doing, because everybody in that trial got one of the two.

The smaller SELF trial measured its patients at three, six and twelve months, although 26 of its 86 patients were left out of the analysis for missing their three-month data, so it is a small trial, not a definitive one (Littlewood and colleagues, 2016). Neither trial set out to answer how long recovery takes. But nobody designs a 12-month follow-up for a problem that clears up in a fortnight.

Recovery is also not always complete. Lewis (2016) notes that some people do not recover fully and that symptoms can return, which is not a comfortable sentence and is better said out loud than discovered at month five.

The two dates you can actually use

Forget predicting the end. Two dates are published, both are decision points, and both can go in a diary on day one. They count from the day treatment starts, not from the day the pain started, so a shoulder that has been sore for months without a program has not banked the time: the diary starts when the program does. If that is you, start a program and set the review date now instead of adding up how long you have already waited.

Week 6 is the first. The UK shoulder surgeons' care pathway runs physiotherapy for about 6 weeks, and says a further 6 weeks is justified if the patient is improving in that first block (Kulkarni and colleagues, 2015).

So the week 6 question is not whether you are better. It is whether anything is moving. A shoulder that has shifted in the right direction earns the next six weeks. One that has not shifted at all earns a different conversation, not a repeat of the same program.

Week 12 is the second, and it is the same number from both documents for two different reasons. It is when a loading program has had long enough to show what it can do (Littlewood and colleagues, 2015), and it is when the 2025 guideline turns to a scan, if symptoms have not resolved or improved, and to a musculoskeletal specialist opinion for someone with severe and persistent pain or disability despite appropriate nonsurgical care (Desmeules and colleagues, 2025). Worth knowing who that means before you go asking for one, because it is not a phrase most people have heard: a primary care sports physician, a specialist in rehabilitation medicine (a physiatrist), or an orthopedic surgeon (Desmeules and colleagues, 2025).

Both dates belong to treatment without surgery. If your cuff has been repaired, your surgeon's own milestones replace them completely, and rehab after rotator cuff repair is the page for those. Neither date is a limit on getting help sooner. They describe the ordinary course of a shoulder that is improving slowly, and they are no kind of instruction to sit out a shoulder that is getting worse, that has become weak, or that has any of the warning signs at the foot of this page.

Four things to agree with your physio at the start, so the months do not just drift: what is being loaded, at what dose, what you are both measuring, and the date you will look at it together. Your physio will set the dose and adjust it as you go.

What gets better first, and what lags

Dates are no use without something to measure against them, so here is the ruler. Night pain usually improves before anything else, then reaching to chest height, then lifting the arm out to the side, then reaching overhead with something in your hand. Two things lag behind the rest: reaching up to a high shelf while holding something, and reaching behind your back. The full version of that order, and why those last two make good personal markers, is on the rotator cuff tendinopathy program.

Use it like this. Pick two everyday movements you cannot do comfortably today, write them down with the date, and look again in a month. Day-to-day pain in a cuff shoulder is noisy enough that a week tells you almost nothing. Check it hourly and you will decide nothing is happening while something is. A month is the right window for a shoulder that is improving slowly or standing still. One that is getting worse, or that has become weak, is not waiting a month for its review.

The nights are the sensitive end of that ruler, and they move in both directions. Improving nights are usually the earliest sign a program is working, and nights that are worse than last week's are usually the earliest sign the load has crept up. That makes them a dose meter, not a diagnosis. Nights that are worse, constant and unrelieved by any position are a different matter and have their own line in the warning list below.

What does healed mean for a tendon?

Not what most people assume, and the gap matters because it decides what you are waiting for.

Every figure on this page is a measurement of pain and function. The trials used questionnaires about what a shoulder can do and how much it hurts, and the 2025 guideline's own answer to the question of timing is to use outcome tools and performance measures rather than dates (Desmeules and colleagues, 2025). Not one of them measures how a tendon looks on a scan.

That distinction is not pedantry, because the two do not track each other. The UK shoulder surgeons' pathway states that a rotator cuff tear on imaging does not always match the symptoms and does not mean an operation is definitely needed, and that a report of a partial-thickness tear is common, with such tears sometimes causing no symptoms at all or being a false positive report (Kulkarni and colleagues, 2015). Australia's healthdirect makes the same point in plainer words: it is common for older people with rotator cuff injuries to have no symptoms. What a scan can and cannot tell you is covered on rotator cuff tear or tendonitis and on do you need a scan for shoulder impingement.

So healed, for this shoulder, is defined by what it does: sleeping through, working overhead, lifting what you need to lift, without the pain coming back afterwards. Waiting for a tendon to look normal on a report is waiting for something nobody has told you is required.

Why is it not getting better?

Three checks before you conclude the program has failed, and a fourth that is not about the program at all.

Check the figure you are judging it against first. If you are at week ten and comparing yourself with a six-week promise from a blog, the program may be perfectly on course. That is the whole point of keeping the three figures apart.

Check the dose next. The best dose for this condition is still not settled, and the 2021 review says so plainly while recommending that programs be individualized and progressive (Boland and colleagues, 2021). The same review found supervised and home-based programs equally effective, which is reassuring if you are doing it at home and also means more appointments are not automatically the missing ingredient. What the dose should look like, and the evidence behind it, is on the rotator cuff tendinopathy program.

Check what else is loading the shoulder. A program doing its job three times a week can be undone by a working day or a training week that keeps the arm above shoulder height, which is where the gym with a rotator cuff injury and exercises to avoid with a rotator cuff problem come in.

And the fourth: a shoulder that is stiff as well as sore, in every direction, including when someone else moves the arm for you, may not be this condition. A frozen shoulder runs on a far longer clock, which is set out on how long frozen shoulder lasts. A shoulder that has become weak, not just sore, is one to get examined rather than reasoned about.

Two structural reasons sit under this heading too. A chalky calcium deposit in the cuff tendon (calcific tendinopathy) behaves differently from a sensitive tendon: the UK shoulder surgeons' pathway describes the acute form as severely painful, often mimicking the pain of a cancer, and as usually needing an early referral for more interventional treatment rather than a longer course of the same exercise (Kulkarni and colleagues, 2015). The other is that the pain may not be coming from the cuff at all. The small joint on top of the shoulder (acromioclavicular joint), the ball and socket joint itself and the neck all put pain into the same area, which is why the 2025 guideline includes screening the neck in a shoulder examination (Desmeules and colleagues, 2025). Both are part of why week 12 turns to a scan instead of another block of what has not worked.

When waiting is the wrong answer

Any page about time has to carry this part, because the harm a timeline does is a specific one: it talks somebody into waiting out a thing that needed seeing today.

The NHS puts the first trigger early. See a doctor or physiotherapist if shoulder pain is getting worse, if it has not improved after 2 weeks, or if it is very difficult to move your arm or shoulder. That is a long way before any of the figures above.

The pain started after a fall, a pull or another injury, especially if the arm has been weak since, for example you cannot lift it out to the side or hold it up. This can be a broken bone or a torn tendon. Get medical advice the same day, and say how it happened and that the arm has been weak since, because UK shoulder surgeons advise urgent referral for a rotator cuff tear caused by an injury.

That one holds however long ago the injury was. An arm that has been weak since a fall two months ago needs examining, not more time.

Sudden, very bad shoulder pain, or you suddenly cannot lift or move your arm at all. Get medical help the same day.

If your cuff has been repaired, one more belongs here. It is the one most likely to get waited out by somebody reading about timelines.

Since your operation, the arm has lost movement or strength it had gained, so lifting or turning it is weaker or much more painful than it was a week or two ago, or you notice clicking, grating or a new catch in the shoulder. A repair can pull off the bone without a pop, and when it does the usual sign is lift that was coming back and has gone again, sometimes with little new pain. Rest the arm, leave out any exercise that makes it worse, contact your surgical team the same day and ask whether the repair needs checking. If you cannot reach them, get urgent medical advice.

One sign on this list has nothing to do with the shoulder and nothing to do with how long anything takes. If chest pain, pressure or tightness does not ease quickly when you rest, spreads to your arm, neck, jaw, stomach or back, or comes with sweating, feeling sick, feeling light-headed or being short of breath, call emergency services straight away, as this can be a heart attack. If it eases quickly and none of these happen, get medical advice the same day, and do not exercise again until you have been checked. If you faint while exercising, call emergency services, even if you feel fine again quickly.

The nights need their own exception spelled out here rather than left in a list, because this page has just told you that bad nights are normal and that they improve first.

Night pain is common with this condition, but get checked within a few days if the pain is constant, no rest or change of position eases it at all, and it keeps getting worse.

New shoulder pain and you have had cancer, now or in the past, or you are losing weight without trying. See your doctor within a day or two and mention it.

The full list at the foot of this page gives every warning sign with how fast to act on it, and it is the part to read before the figures, not after them. None of it means the months ahead are wasted. It means a few specific things are judged in hours and days rather than in weeks, and the clock on this page is only for the shoulder that none of them applies to.

One sign in the list below appears three times, at three different speeds. Pain at night or at rest that is constant, that no rest or change of position eases at all, and that keeps getting worse is a same-day matter if you are being treated for cancer now, a within-a-day-or-two matter if you have had cancer at any time or are losing weight without trying, and otherwise needs checking within a few days. If you cannot tell which of those you have, treat it as the more urgent one.

Related exercise programs

• Rotator cuff tendinopathy
• Shoulder impingement
• Rotator cuff tear without surgery
• Rotator cuff repair rehab

Common questions

References

1. Desmeules F, Roy JS, Lafrance S, et al. Rotator cuff tendinopathy diagnosis, nonsurgical medical care, and rehabilitation: a clinical practice guideline. Journal of Orthopaedic and Sports Physical Therapy. 2025;55(4):235-274. https://doi.org/10.2519/jospt.2025.13182
2. Littlewood C, Malliaras P, Chance-Larsen K. Therapeutic exercise for rotator cuff tendinopathy: a systematic review of contextual factors and prescription parameters. International Journal of Rehabilitation Research. 2015;38(2):95-106. https://doi.org/10.1097/MRR.0000000000000113
3. Braun C, Hanchard NC, Batterham AM, Handoll HH, Betthauser A. Prognostic models in adults undergoing physical therapy for rotator cuff disorders: systematic review. Physical Therapy. 2016;96(7):961-971. https://doi.org/10.2522/ptj.20150475
4. Boland K, Smith C, Bond H, Briggs S, Walton J. Current concepts in the rehabilitation of rotator cuff related disorders. Journal of Clinical Orthopaedics and Trauma. 2021;18:13-19. https://doi.org/10.1016/j.jcot.2021.04.007
5. Kulkarni R, Gibson J, Brownson P, Thomas M, Rangan A, Carr AJ, Rees JL. Subacromial shoulder pain. BESS/BOA Patient Care Pathways. Shoulder and Elbow. 2015;7(2):135-143. https://doi.org/10.1177/1758573215576456
6. Hopewell S, Keene DJ, Marian IR, et al. Progressive exercise compared with best practice advice, with or without corticosteroid injection, for the treatment of patients with rotator cuff disorders (GRASP): a multicentre, pragmatic, 2 x 2 factorial, randomised controlled trial. The Lancet. 2021;398(10298):416-428. https://doi.org/10.1016/S0140-6736(21)00846-1
7. Littlewood C, Bateman M, Brown K, Bury J, Mawson S, May S, Walters SJ. A self-managed single exercise programme versus usual physiotherapy treatment for rotator cuff tendinopathy: a randomised controlled trial (the SELF study). Clinical Rehabilitation. 2016;30(7):686-696. https://doi.org/10.1177/0269215515593784
8. Gutman MJ, Joyce CD, Patel MS, et al. Early repair of traumatic rotator cuff tears improves functional outcomes. Journal of Shoulder and Elbow Surgery. 2021;30(11):2475-2483. https://doi.org/10.1016/j.jse.2021.03.134
9. Bjornsson HC, Norlin R, Johansson K, Adolfsson LE. The influence of age, delay of repair, and tendon involvement in acute rotator cuff tears: structural and clinical outcomes after repair of 42 shoulders. Acta Orthopaedica. 2011;82(2):187-192. https://doi.org/10.3109/17453674.2011.566144
10. Lewis J. Rotator cuff related shoulder pain: assessment, management and uncertainties. Manual Therapy. 2016;23:57-68. https://doi.org/10.1016/j.math.2016.03.009
11. NHS. Shoulder pain. Page last reviewed 22 May 2023. https://www.nhs.uk/symptoms/shoulder-pain/
12. American Academy of Orthopaedic Surgeons. Shoulder impingement/rotator cuff tendinitis. OrthoInfo. https://www.orthoinfo.org/diseases--conditions/shoulder-impingementrotator-cuff-tendinitis/
13. healthdirect. Rotator cuff injury. https://www.healthdirect.gov.au/accidents-and-injuries/rotator-cuff-injury
14. Healthify He Puna Waiora. Shoulder pain. https://healthify.nz/health-a-z/s/shoulder-pain
15. NHS. Heart attack. Page last reviewed 31 March 2026. https://www.nhs.uk/conditions/heart-attack/
16. NHS. Dislocated shoulder. Page last reviewed 16 September 2026. https://www.nhs.uk/conditions/dislocated-shoulder/
17. NHS. Polymyalgia rheumatica. Page last reviewed 23 April 2023. https://www.nhs.uk/conditions/polymyalgia-rheumatica/
18. NHS. Temporal arteritis. Page last reviewed 14 August 2023. https://www.nhs.uk/conditions/temporal-arteritis/

Written and checked by the PocketPhysio editorial team. Last updated 2026-11-05.

This page is general education, not a diagnosis or a personal treatment plan. See a physiotherapist or doctor for advice about your own situation.