Shoulder internal rotation: it is the movement you use every time you reach your hand behind your back. Fastening a bra, threading a belt through the loops, taking a wallet out of a back pocket, scratching between your shoulder blades. When it is reduced you notice it through these gestures long before you notice it in a test.
Let me tell you up front how this article is structured, because it is different from almost all the others you will find on the topic. The standard story says that reduced internal rotation leads to compensations, then to impingement, then to rotator cuff injuries, and that stretching the posterior capsule fixes everything. When I went to check what really holds up in the literature, part of that chain stands and part does not. You will find both below, and above all how to use these exercises anyway, knowing what to expect.
⚠ Important: read this before continuing
The exercises and advice in this article are general indications and may not suit your situation. Every person has a unique clinical history: what works for others could be counterproductive for you.
Want a professional's opinion? Start with the questionnaire: you register and see who will follow you, a physiotherapist registered with the Register chosen for your case, to whom you write in chat. If you would rather hear their voice, the first consultation is free and we look at your situation together.
Before stretching: when it is not simple stiffness
A shoulder that rotates poorly is almost always a benign matter. Some situations, however, are not treated with stretching: they need to be looked at. See a doctor, without delay, if you recognise any of these.
- You have also lost external rotation, and you lose it even when someone else moves your arm while you stay relaxed. Loss of movement in several directions, and not only inwards, is the typical picture of adhesive capsulitis and glenohumeral osteoarthritis: these follow a different pathway from the one described here.
- The stiffness appeared after a trauma: a fall onto the arm, a dislocation, a direct blow. All the more so if the arm can no longer be lifted actively or you can see a deformity of the outline of the shoulder.
- Night pain that never changes with position and wakes you systematically, especially if accompanied by fever, sweating, unexplained weight loss or a personal history of cancer.
- Tingling or real weakness affecting the whole hand or the whole arm, with difficulty in fine movements. Here the suspicion shifts to the neck and the nerves, and shoulder stretching has nothing to do with it.
- Left shoulder pain together with chest tightness, shortness of breath or cold sweating: this is a medical emergency, call 112. Likewise, right shoulder pain that appears together with upper abdominal pain after meals should be assessed.
None of these lists is a diagnostic test. They are reasons to get checked, not reasons to panic. A 2024 systematic review comparing 26 clinical practice guidelines on shoulder disorders finds exactly here one of the few points of agreement: the initial assessment of shoulder pain starts from the clinical history, the search for red flags and the clinical examination, while magnetic resonance imaging is usually not recommended in the early stages.
Shoulder internal rotation: what it is, in practical terms
Shoulder internal rotation is the rotation of the humerus inwards around its own axis. We measure it in two different positions, and they are not interchangeable: with the arm along the body, bringing the hand behind the back, or with the arm abducted to 90 degrees and the elbow bent, lowering the forearm towards the couch. It is this second position, with the scapula held still, that is used in the studies.
The muscles that produce it are the subscapularis, the latissimus dorsi, the pectoralis major and the teres major. But when the movement is missing, the brake is almost never muscular: it is the back part of the joint capsule and the tissues lying over it, which become less extensible. This is why the classic intervention is called stretching of the posterior capsule and not "stretching of the internal rotators".
One detail that changes expectations: in people who have thrown for years, and who started as children, part of the lost rotation is not stiff tissue but bone. A 2020 systematic review in Sports Health, which gathered 32 studies on baseball players, found that humeral torsion explains a large part of the internal rotation available, since for every extra degree of internal rotation there were 0.65 extra degrees of torsion, while it did not explain external rotation or horizontal adduction. The authors' conclusion is that much of internal rotation may be non-modifiable: it is the shape of the bone, and there is nothing to stretch. This is one of the reasons why the gains in the studies are more modest than one would expect, and when you read that shoulder internal rotation can be "unlocked", this is the physical limit that nobody talks about.
The same review adds a finding that cools enthusiasm even further: between injured and uninjured players, humeral torsion did not differ. Bone adaptation, in itself, was not found to be linked to the risk of getting hurt.
Why it gets reduced
The routes are few and fairly recognisable.
Repeated overhead movement. Baseball, volleyball, tennis, water polo, javelin throwing. Repeating the deceleration of the arm loads the back of the shoulder heavily, which gradually stiffens. In this population the loss of internal rotation is so common that it has a name of its own in the literature, GIRD, short for glenohumeral internal rotation deficit.
Prolonged immobility. After a sling, after an operation, or simply after weeks in which the arm stopped going behind the back because it hurt. The capsule shortens where it is not taken.
The pain itself. A shoulder that hurts spontaneously limits its own movement, and that limitation is a consequence before it is a cause. This is an important point when you decide what to treat first.
Positions held for a long time. Hours with the arms reaching forward on a keyboard shorten the front of the chest and change how the scapula rests. Here, however, we are in the territory of clinical observation, not of numbers: none of the studies I cite below tested shoulder internal rotation in sedentary people.
The link with injuries: what holds up and what does not
This is the part where the article you are reading separates from the others, and it is worth being precise. The available reviews all point in the same direction: people with the deficit get injured a little more often, but they do not agree on how solid that signal is: two reach different conclusions on the same years, and the difference lies in the design of the studies they included.
In 2018, in Sports Health, a systematic review with meta-analysis gathered 17 studies, 10 of which had usable movement measures, for a total of 2,195 athletes (1,889 men and 306 women, mean age 20.8 years). The result: shoulders with an internal rotation deficit tended towards injury, with a mean difference of 3.11 degrees, but with a confidence interval from −0.13 to 6.36 and a p value of 0.06. The same for total arc of motion and for external rotation gain. The authors' conclusions are unambiguous: none of the shoulder movement measures reached statistical significance in the correlation with shoulder or elbow injury.
Another meta-analysis from the same year, in the Orthopaedic Journal of Sports Medicine, points the same way but this time crosses the threshold of significance: across 819 overhead athletes (226 injured and 593 not), the standardised mean difference between the two groups was 0.46 with a confidence interval from 0.15 to 0.77. The weighted mean deficit was 13.8 degrees in the injured against 9.6 in the uninjured. The effect, however, was clear in adults (0.60, from 0.18 to 1.02) and not in adolescents (0.20, from −0.24 to 0.63), the heterogeneity between studies was moderate, and the authors close by calling for better-quality prospective research to clarify the role of the deficit in future injuries.
That prospective research does exist, and it is the most solid voice in the group. Also in 2018, in the Journal of Athletic Training, a review selected only studies that measured movement in healthy baseball players and then followed them over time to record injuries. Six articles, 3 of them in the meta-analysis, with methodological quality from moderate to high. There the internal rotation deficit was predictive of injury (4.28, confidence interval from 0.71 to 7.86, p = 0.02), together with loss of total arc of motion (−6.19, from −10.28 to −2.10, p = 0.003) and horizontal adduction deficit (−8.32, from −12.08 to −4.56, p < 0.001). External rotation, by contrast, predicted nothing.
How should this body of evidence be read honestly? Like this: in throwers, measuring internal rotation and total arc makes sense and gives information about risk. Outside that context, if you are someone who sits in an office and cannot fasten your bra, that body of evidence simply does not speak about you. And even within that context, the association is not as clear-cut as the headlines suggest.
An important tip: before trying these exercises, make sure you have received a correct diagnosis. People often think they have one problem, but the real cause of the pain is something else. Doing the wrong exercises can make things worse instead of better.
How to assess yourself
The absolute number matters little, the comparison between the two sides matters a lot. Put one hand behind your back, your back resting against the spine, and try to climb as high as you can with your thumb: take as a reference the vertebra or the anatomical point you reach. Then repeat on the other side. What you are interested in is the difference.
A second useful check is horizontal adduction: lying on your back, bring the straight arm across the chest towards the opposite shoulder, keeping the shoulder you are stretching resting on the floor. Again, compare the two sides. In the prospective studies on baseball, this deficit was the measure with the widest numerical gap between those who got injured and those who did not, although the authors note that the results on this item were less uniform across studies and that more research is needed before considering it a safe target.
If you notice a marked difference and you also have pain, it is time to get assessed instead of stretching blindly. If the movement is difficult but not painful, the exercises below are a reasonable starting point.
Exercises for shoulder internal rotation, in order of how well documented they are
Cross-body stretch
It is the most studied and the one with the best numbers. Standing or seated, bring the straight arm across the chest towards the opposite shoulder and guide it at the elbow with the other hand. The technical point that makes the difference: do not let the shoulder blade slide forwards. If the shoulder rotates as one block, you are stretching very little.
In the reference randomised study, published in 2007 in the Journal of Orthopaedic & Sports Physical Therapy, 54 asymptomatic subjects were followed for 4 weeks, with the stretching performed on their own, once a day, 5 repetitions of 30 seconds, on the more limited side only. The cross-body group gained on average 20.0 degrees of internal rotation against 5.9 for the control group (p = 0.009).
It is worth knowing how that comparison was built, because it is usually not reported. The control group was not people with a stiff shoulder left untreated: they were the 24 participants who had less than 10 degrees of difference between the two sides, that is, people without the problem. Those with more than 10 degrees of difference were randomised between sleeper and cross-body, 15 per group. So the "20 against 5.9" compares people stretching a limited shoulder with people who had nothing to recover, and the authors themselves write the conclusion in these terms. It is a result in favour of the cross-body, not the clean proof that it appears to be at first sight.
Sleeper stretch
Lying on your side on the side to be treated, shoulder flexed to 90 degrees, elbow bent: with the other hand you slowly push the forearm towards the couch. The idea is to lock the scapula with the weight of the body in order to isolate the joint.
Here I have to give you the uncomfortable news, because it is the most recommended exercise of all and the evidence is more uncertain than it is made out to be. In the same 2007 study, the sleeper stretch gained 12.4 degrees, a value that was not significant compared with that control group (p = 0.586). It must be said in full, however: not even the difference between sleeper and cross-body was significant (p = 0.148), and the authors themselves write that the small sample probably prevented detecting a difference between the two groups, which was nonetheless potentially relevant in practice. Put plainly: that study does not prove that the sleeper does not work, it shows that it was too small to say. A 2017 systematic review in the Journal of Sport Rehabilitation, which gathered 10 randomised studies, reaches a consistent conclusion: moderate evidence in favour of the cross-body in the short term in young asymptomatic people, and moderate evidence that the active sleeper stretch may be no more effective than doing nothing.
The most recent data rebalances the scales. A meta-analysis published in PM&R in 2025 included 6 randomised studies on 255 overhead athletes with a documented deficit: the sleeper stretch, alone or combined with the cross-body, produced an average gain of 7 degrees of internal rotation compared with control (confidence interval from 1 to 13, p = 0.02) and of 6 degrees of horizontal adduction.
Seven degrees. It is a real and measurable gain, and it is also the measure of how inflated some promises are.
On how to perform it: stop before a tingling appears in the front or deep part of the shoulder. That tingling is not the stretch working, it is the shoulder defending itself, and continuing to push is the most common way of making an already irritated shoulder worse.
Thoracic mobility
It does not act on the rotation itself but on what lies beneath it. Seated or on all fours, rotate the trunk, bringing one arm towards the ceiling and following the hand with your gaze. A thorax that rotates little forces the shoulder to look for degrees it does not have. I have no numbers to give you on this specific exercise and I will not invent any.
Pectoral opening at the doorway
Forearm resting on the doorframe, elbow at shoulder height, and you move forward half a step. Useful if you work many hours with your arms forward. Here too: sensible, not demonstrated for this purpose.
Pressure on the back of the shoulder
A ball between the shoulder blade and the wall, slowly searching for the sensitive points of the back of the area. It is low-risk and gives immediate relief to many people, but it should be used as a warm-up before moving, not as a treatment.
Strengthening, which is the part that gets forgotten
Gaining mobility without the control to manage it is a poor deal. Sensible work runs on three fronts: the external rotators with a band, with the elbow resting at the side and slow movements; the muscles that govern the scapula, because a stable scapula lets the humerus move better; and endurance, not only maximal strength, since shoulder problems almost always arrive at the end of a long set or a long session. If you want a complete programme, I have described it in the shoulder strengthening exercises and, for the specific part on the tendons, in the rotator cuff guide.
The question nobody asks: does stretching really prevent injuries?
Here we reach the most delicate point, and it is where the current story makes a logical jump. We know that stretching gains degrees. We know that in throwers a deficit is associated with more injuries. But the conclusion, that stretching therefore prevents injuries, is a link that has to be demonstrated separately, and to date it has not been.
A 2021 systematic review in Physical Therapy in Sport looked for all the randomised or prospective studies that applied shoulder prevention programmes in overhead athletes and measured their injury incidence. It found 7. Two reported a reduction in injuries with strength and flexibility exercises (odds ratio 0.72, confidence interval from 0.52 to 0.98, and odds ratio 0.22, from 0.06 to 0.75). One reported a reduction with stretching of the posterior capsule (hazard ratio 0.36, from 0.13 to 0.95). These are encouraging numbers, and it would be wrong to hide them.
But the authors' conclusion is the part nobody quotes: only 3 of the 7 studies showed a favourable effect, only one was at low risk of bias, and at present no conclusions can be drawn about the effectiveness of shoulder prevention programmes in the overhead athlete.
How I translate this for you: stretching the posterior capsule is a reasonable bet, low-risk and with a plausible mechanism behind it. It is not an insurance policy. If you are doing five minutes of sleeper stretch a day convinced it is protecting you, know that this protection is a hypothesis, not a result.
What to expect, realistically
The studies I have cited measured gains in shoulder internal rotation over four weeks of daily work, and they produced results in the order of a few degrees on average. Not days, not miracles, and not all the degrees you are missing, especially if part of the limitation is bone adaptation from years of throwing.
The mistakes I see most often are three. Pushing into pain, believing that intensity speeds up the result: it does not, and on an irritated shoulder it is the quickest way to have to stop. Stretching only the limited side without ever strengthening: you recover movement that nobody controls. And stretching a shoulder that actually has a different problem (a capsulitis that is also taking away external rotation, or a pain that starts in the neck), and then wondering after two months why nothing changes. If the picture does not add up for you, the article on when to worry about shoulder pain will help you frame it.
In practice
An honest summary of everything above. Shoulder internal rotation can be improved: randomised studies show it, with measurable but modest gains, and the cross-body stretch has evidence at least equal to that of the sleeper stretch, the exercise that everyone recommends first. The link between deficit and injury is documented in throwers, in a less clear-cut way than it is told, and practically unstudied in people who do not play overhead sports. That recovering degrees prevents injuries remains a plausible hypothesis and not a demonstrated one.
This is not a reason to do nothing. It is a reason to do things with the right expectations: stretch without pushing into pain, always compare the two sides, put shoulder and scapula strengthening alongside the mobility work, and if the signs listed at the beginning are present, get checked before you start. If you have limitations that block an everyday gesture or your sport, Ri-Hub's physiotherapists can assess you and build the programme around your situation, all comfortably online.
You might also be interested in:
Myofascial Syndrome and Trigger Points: A Complete Guide to Remedies Carpal Tunnel Exercises: 10 Effective Movements to Reduce Symptoms Neck Pain and a Numb Arm: Causes and Remedies
Stay in touch with us Join our dedicated Facebook group: we answer all patients, give practical advice and clear up the most common doubts.
Sources
- Glenohumeral Internal Rotation Deficit and Risk of Upper Extremity Injury in Overhead Athletes: A Meta-Analysis and Systematic Review, Sports Health, 2018
- Glenohumeral Internal Rotation Deficit and Injuries: A Systematic Review and Meta-analysis, Orthopaedic Journal of Sports Medicine, 2018
- Shoulder Range of Motion and Baseball Arm Injuries: A Systematic Review and Meta-Analysis, Journal of Athletic Training, 2018
- The Relationship Between Humeral Torsion and Arm Injury in Baseball Players: A Systematic Review and Meta-analysis, Sports Health, 2020
- Role of stretching interventions in enhancing the shoulder range of motion in overhead athletes with glenohumeral internal rotation deficit: A systematic review and meta-analysis, PM&R, published online August 2025 (issue 2026;18(2):200-209)
- Effectiveness of Stretching on Posterior Shoulder Tightness and Glenohumeral Internal-Rotation Deficit: A Systematic Review of Randomized Controlled Trials, Journal of Sport Rehabilitation, 2017
- A randomized controlled comparison of stretching procedures for posterior shoulder tightness, Journal of Orthopaedic & Sports Physical Therapy, 2007
- Effectiveness of shoulder injury prevention programs in an overhead athletic population: A systematic review, Physical Therapy in Sport, 2021
- A Systematic Review of Clinical Practice Guidelines on the Diagnosis and Management of Various Shoulder Disorders, Archives of Physical Medicine and Rehabilitation, 2024
Last source check: August 2026.
Need a consultation?
Start with the questionnaire on Ri-Hub, a few questions about your case, you register, and you see who will follow you: a physiotherapist registered with the Register, whom you can write to in chat whenever you like. If you would rather talk by voice, the first consultation is free, lasts twenty minutes and commits you to nothing: they arrive having already read what you wrote, and together you decide whether and how to go on.
