Does shoulder impingement show on an MRI, and what would a scan change?

Does shoulder impingement show on an MRI?

Partly, and the part it misses is the part you are asking about. An MRI gives a good view of the rotator cuff tendons and the soft tissue around them, it can show fluid or inflammation in the bursa or the cuff, and in some cases it shows tears in the tendons (American Academy of Orthopaedic Surgeons). The tissues that hurt in this condition are all visible.

What is not visible is impingement itself. The word describes something that happens while you move: pain as the arm comes up, explained in the old model by a tendon being squeezed under the bony tip of the shoulder. A scanner images a shoulder that is lying still. Both Dutch Orthopaedic Association guidelines build the diagnosis out of a physical examination instead, and the 2026 update states it in one line, to perform a cluster of physical diagnostic tests to diagnose subacromial pain syndrome (Lambers Heerspink and colleagues, 2026).

That name is worth knowing, because it is likely to be the one on your report. The same guideline defines subacromial pain syndrome as non-traumatic, non-rheumatologic shoulder pain that is particularly painful during arm elevation, and gathers supraspinatus tendinosis, calcific tendinitis and degenerative supraspinatus tears inside it. Our rotator cuff tendinopathy page is the same shoulder under a third name, and the shoulder impingement program explains how the names came to move while the shoulder stayed the same.

So when a scan is requested here, the question it carries is narrower than yours. The 2014 guideline recommends ultrasound to exclude a rotator cuff rupture, and the 2026 update asks for ultrasound where a full-thickness or partial-thickness rupture of the supraspinatus tendon is suspected, with MRI considered if ultrasound is not available or comes back inconclusive. The radiologist is being asked whether a tendon is torn. Nobody is being asked to confirm a label.

Does an X-ray show shoulder impingement?

No. X-rays do not show the soft tissues of the shoulder such as the rotator cuff, so a plain X-ray of a shoulder with cuff pain is usually normal, or may show a small bone spur, and a particular projection called the outlet view sometimes shows a small spur on the front edge of the acromion (American Academy of Orthopaedic Surgeons). A spur on a film and a cause of pain are not the same thing. The 2014 Dutch guideline lists limiting imaging to conventional radiographic examination among the things clinicians should not do (Diercks and colleagues, 2014).

X-rays still have work to do in a sore shoulder, and most of it is exclusion. The 2014 guideline pairs imaging with conventional radiography of the shoulder to look for osteoarthritis. The BESS and BOA pathway, written by UK shoulder surgeons, puts it like this: although shoulder X-rays with two views in primary care can be useful in patients not improving with conservative treatment, imaging of the rotator cuff with ultrasound or MRI is rarely indicated in primary care (Kulkarni and colleagues, 2015).

Two neighboring problems show themselves on a plain film more readily than this one does. A calcium deposit in a tendon turns up on an X-ray or an ultrasound, and calcific tendinitis covers what that means. An X-ray is also used to rule out arthritis in a shoulder that has gone genuinely stiff, which is frozen shoulder territory, not this page's.

Why the scan rarely changes the treatment

Because a finding is not the same as a cause, and the surgeons' pathway says so plainly. It notes that partial thickness tears are a common ultrasound finding and that such tears can be asymptomatic or the consequence of false positive reporting. It warns that ultrasound is user dependent and that the accuracy of ultrasound in identifying rotator cuff tears varies. And it states that the presence of a rotator cuff tear on imaging does not always correlate with symptoms and does not imply a definite need for surgery in every case (Kulkarni and colleagues, 2015).

Follow that through and one line in the pathway carries more weight than any report. Referrals to secondary care, it says, should be based on patient symptoms, disability and response to conservative treatment. Not on the images.

A 2025 clinical practice guideline written for physical therapists lands in the same place. Desmeules and colleagues (2025) advise clinicians not to order an imaging test to confirm rotator cuff tendinopathy in the first management of an adult with shoulder pain, and say imaging may be considered if symptoms have not resolved or improved within a maximum of 12 weeks of appropriate nonsurgical care. Both of those sit at that guideline's grade F, which it defines as the expert opinion of the guideline group rather than a finding from studies, so read them as what experienced clinicians agreed on rather than as something trials have settled. The 12 weeks is counted from the start of treatment, which is a different clock from the 2014 Dutch guideline's 6 weeks of symptoms.

The other reason is less about the image and more about what happens next. The 2014 guideline says subacromial pain syndrome should preferably be treated non-operatively and that there is no convincing evidence that surgical treatment is more effective than conservative management (Diercks and colleagues, 2014). The first treatment is a graded exercise program either way, which is set out on the shoulder impingement program and in more tendon detail on the rotator cuff tendinopathy program. A report cannot tell you how much that arm will be doing in six weeks, and that is usually what people came to find out. What causes shoulder impingement deals with the mechanism question behind it.

When does a scan actually change what happens next?

Five situations. They are worth naming, because none of what follows adds up to imaging being useless.

1. A torn tendon is suspected. This is the main one. The 2026 update asks for ultrasound where a rupture of the supraspinatus tendon is suspected, to guide treatment decisions, with MRI if the ultrasound is unavailable or inconclusive. An injury that has left the arm weak is the urgent version: UK shoulder surgeons write that an acute cuff tear as a result of a traumatic event needs urgent referral and ideally should be seen in the next available outpatient clinic (Kulkarni and colleagues, 2015). If that is your story, act on the first warning sign at the foot of this page today instead of waiting for an imaging appointment, and impingement, bursitis or a rotator cuff tear goes into the difference between those three.
2. Things are not settling. The 2014 guideline's own summary says diagnostic imaging is useful after 6 weeks of symptoms, and the surgeons' pathway puts two-view X-rays in primary care for patients who are not improving with nonsurgical treatment.
3. A specialist is now involved. The pathway suggests considering imaging with ultrasound or MRI to assess the integrity of the rotator cuff muscles and tendons, and says this is more usefully performed after referral to secondary care, where it can help direct treatment there.
4. An operation is being planned. Here MRI has something ultrasound cannot match: the 2026 update says MRI is superior to ultrasound for assessing rotator cuff atrophy and fatty infiltration, and reserves it for inconclusive ultrasound or for the detailed assessment of muscle needed in surgical planning.
5. Something does not fit the pattern. Pain that spreads below the elbow, pins and needles, a shoulder that will not move in any direction, or anything in the warning signs below, all raise other possibilities, and imaging is one way of chasing them.

The decision itself is not yours to make alone and not this page's to make for you. It belongs to the clinician who has examined your shoulder, and the useful thing you can bring to that conversation is in the last two sections.

What the words in a scan report mean

Plain English for the vocabulary, and that is all it is. Nothing below reads your report for you, and no web page can do that job, because each of these words has to be weighed against the shoulder it came from.

Tendinosis, or increased signal in a tendon, says the tissue has altered in make-up and appearance. It is what tendinopathy looks like to a scanner. The 2026 Dutch guideline counts supraspinatus tendinosis inside subacromial pain syndrome, so in this condition it is an expected finding rather than a shock, and it is not the same as a tear.

Bursal fluid, or bursitis, means fluid in the lubricating sac that sits between the rotator cuff and the bone on top of the shoulder blade (American Academy of Orthopaedic Surgeons). An MRI can show fluid or inflammation in that bursa or in the cuff itself. This is the word that frightens people most and settles least.

A partial thickness tear is one that goes some of the way through the tendon and leaves the remainder joined. Remember what the surgeons' pathway says about this particular finding on ultrasound: it is common, and it can be symptomless or a false positive report (Kulkarni and colleagues, 2015).

A full thickness tear goes from one surface of the tendon to the other where it has torn. A report may then add how large it is and how far it has pulled back, the sort of detail the 2014 guideline says MRI contributes, along with the matching muscle atrophy and fatty infiltration.

Atrophy and fatty infiltration describe muscle that has thinned or taken on fat in place of muscle fiber. It usually speaks to a problem of long standing, and it is what the 2026 update says MRI judges better than ultrasound when surgery is being planned.

A subacromial spur, or a comment on the shape of the acromion, refers to extra bone on the underside or front edge of the bony tip of the shoulder. Our shoulder impingement program covers what the trials found when surgeons shaved that bone away. Plain words for the anatomy are in the glossary.

If you are holding a report with any of these words in it, the next step is the clinician who examined you, not a search engine. Ask what it changes. There is a fair chance the answer is nothing, and nothing is a better answer than it sounds.

If the scan shows a tear, does that mean an operation?

Not by itself, and two guidelines say so from different directions. The surgeons' pathway states that a rotator cuff tear on imaging does not always correlate with symptoms and does not imply a definite need for surgery in every case (Kulkarni and colleagues, 2015). The 2014 Dutch guideline is blunter about one group: there is no indication for the surgical treatment of asymptomatic rotator cuff tears (Diercks and colleagues, 2014).

The exception is the one already named above. A cuff tear caused by a traumatic event needs urgent referral, and if your pain started with a fall or a pull and the arm has been weak since, that is a same-day matter whatever any scan shows or has not shown yet.

Beyond that, this page describes no operation, weighs none against another, and gives no recovery times, because that is a surgeon's conversation held with your shoulder in the room. If an operation has already happened, rotator cuff repair rehab is the page you want, and your surgeon's plan outranks anything written here. What the evidence does support saying is that a tear on a report is information, not an instruction.

How is shoulder impingement diagnosed without a scan?

From your story and an examination, which is how the guidelines ask for it to be done. The 2014 Dutch guideline found that no single test is sufficiently accurate to diagnose subacromial pain syndrome, that a combination of tests increases the post-test probability of the diagnosis, and it recommends a combination of the Hawkins-Kennedy test, the painful arc test and the infraspinatus muscle strength test, with the drop-arm test and infraspinatus and supraspinatus strength testing used where a rotator cuff tear is the question (Diercks and colleagues, 2014).

Those are clinicians' tests, performed and weighed up by someone who is also looking at everything else. They are not a way to label your own shoulder at home, and no single one of them settles anything, which is the guideline's own point. The 2026 update reaches the same place, asking for a cluster of tests rather than one, for a thorough history and a complete physical examination to rule out other causes of shoulder pain, and for multiple tests partly in order to rule out other problems such as rotator cuff tears.

What you can usefully bring is the part only you know: when it started and whether anything was injured, which movements and which heights hurt, and whether the pain sits in a band partway up or everywhere. Say whether it wakes you, and what you could do three months ago that you cannot do now. Say too what you have already tried, for how long, and what happened when you did.

Two things get sorted out at that appointment rather than on a scan. A shoulder that stops short in every direction, however it is moved, is a stiff shoulder, not a painful one, and belongs with frozen shoulder. And the neck gets examined too, to make sure the pain is not coming from a pinched nerve there (American Academy of Orthopaedic Surgeons), which is what the neck pain program and cervical radiculopathy are for. A 2025 physical therapy guideline covers the diagnosis, nonsurgical care and rehabilitation of rotator cuff tendinopathy, including tendinopathy with or without calcification and partial-thickness tears, although it leaves full-thickness tears out of its scope (Desmeules and colleagues, 2025). Our rotator cuff tendinopathy page carries the rest of its position on imaging.

What to ask if you are offered a scan, or refused one

Ask what the scan is for, not whether you can have it. These are the questions that change the answer you get.

1. What question is this scan meant to settle?
2. What would you do differently if it comes back normal, and what would you do differently if it shows a tear?
3. If no scan is planned, what is the working diagnosis, what should be improving, and by when?
4. What happens if it has not improved by then: is the diagnosis reviewed, the program changed, or the scan arranged?
5. Who will explain the report to me, and when?

If you have been told you do not need imaging, the pathway above is the reason, and it is a defensible one, not a brush-off: cuff imaging is rarely indicated in primary care, and the referral decision rests on your symptoms, your disability and how you respond to treatment (Kulkarni and colleagues, 2015). The useful follow-up question is what counts as a fair trial of treatment and when it gets reviewed. If the shoulder is getting worse, has not shifted in 2 weeks, or makes the arm very hard to move, the NHS advice is to see a doctor or physio. A report that nobody has talked you through is worth an appointment of its own; take it with you and ask what it changes.

The same worry turns up at other joints. Do you need a scan for outer hip pain answers it for the hip, and shoulder conditions lists the rest of the shoulder pages.

When shoulder pain needs looking at sooner, whatever the scan says

A normal scan is not a clean bill of health, and that comes from the surgeons' pathway, not from our own caution: a normal rotator cuff ultrasound does not exclude serious shoulder problems such as a tumor or arthritis in the shoulder joint itself, and a suspected tumor or malignancy needs urgent referral through the local 2-week cancer referral pathway (Kulkarni and colleagues, 2015).

So none of the signs below waits for an imaging appointment, and none of them is undone by a scan that came back clear. They are listed under the warning signs at the foot of this page, each with its own timing, from the ones that mean calling emergency services now to the ones that mean an appointment in the next few days.

Related exercise programs

• Shoulder impingement
• Rotator cuff tendinopathy
• Frozen shoulder
• Calcific tendinitis

Common questions

References

1. Diercks R, Bron C, Dorrestijn O, et al. Guideline for diagnosis and treatment of subacromial pain syndrome: a multidisciplinary review by the Dutch Orthopaedic Association. Acta Orthopaedica. 2014;85(3):314-322. https://doi.org/10.3109/17453674.2014.920991
2. Lambers Heerspink FO, Veen EJD, Dorrestijn O, et al. Update of guideline for diagnosis and treatment of subacromial pain syndrome: a multidisciplinary review by the Dutch Orthopedic Association. Part 1: preventive measures, diagnostics, and non-surgical treatment of subacromial pain syndrome. Acta Orthopaedica. 2026;97:91-98. https://doi.org/10.2340/17453674.2026.45365
3. 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
4. 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
5. American Academy of Orthopaedic Surgeons. Shoulder impingement/rotator cuff tendinitis. OrthoInfo. https://www.orthoinfo.org/en/diseases--conditions/shoulder-impingementrotator-cuff-tendinitis/
6. NHS. Shoulder pain. Page last reviewed 22 May 2023. https://www.nhs.uk/symptoms/shoulder-pain/

Written and checked by the PocketPhysio editorial team. Last updated 2026-10-18.

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