Exercises after rotator cuff repair surgery: the sling, the recovery stages and the warning signs

What the operation does

The torn tendon, most often the supraspinatus on top of the shoulder, is sewn back onto the bone with small anchors that carry stitches. It is usually keyhole surgery under a general anaesthetic, often as a day case, with a nerve block that can leave the arm numb and heavy for up to 2 days. Sometimes the tear is too big, or the tendon too fragile, to repair fully, and only part of it can be reattached.

Open, mini-open and keyhole repairs give similar results, according to the American Academy of Orthopaedic Surgeons (AAOS). In the UK Rotator Cuff Surgery trial (UKUFF), 273 people aged 50 and over at 19 NHS hospitals had an open or a keyhole repair, and their shoulder scores were the same at 2 years. Other procedures are often done at the same time, such as fixing the biceps tendon (a biceps tenodesis), which brings its own limits: one NHS plan avoids bending the elbow against resistance for 8 weeks.

Sources 1, 2, 3, 4, 5

What recovery usually looks like

One NHS hospital calls it a good operation for pain relief that is less reliable at improving movement, and says it won't restore normal strength to the arm. It is "not a quick fix": most people notice an improvement by 3 to 6 months, the shoulder feels almost normal by 9 to 12 months, and it keeps improving for up to 18 months, with strength improving for up to 2 years. The AAOS says most people have useful movement and enough strength by 4 to 6 months.

In the UKUFF trial, shoulder scores rose from about 26 to about 42 out of 48 by 2 years, and 85% of people said their shoulder was better. The same NHS hospital says about 1 in 5 people still have some symptoms, and most never get completely normal strength above shoulder height.

Sources 1, 3, 4

Re-tears are commonStrong evidence

A repaired tendon doesn't always heal fully onto the bone. Across 31 higher-quality studies, scans found the tendon had come away again in about 15% to 21%; across 30 studies of 13,161 people, it ranged from 5% to 43%. In a trial of 206 people, a full-thickness tear was there at a year in 30% of one group and 33% of the other. In UKUFF, in people aged 50 and over, it was 39% to 46%.

The risk rises with age and with the size of the tear: it roughly doubles from about 15% at 50 to over 30% at 70. In UKUFF's scans, 66% of small and 68% of medium tears had healed at a year, against 47% of large and 27% of massive ones.

Sources 6, 7, 8, 4, 9, 10

What a re-tear on a scan does and doesn't meanEmerging evidence

People with a re-tear score lower on average, by about 8 points on a 100-point shoulder score, and are weaker lifting the arm out to the side, but in a review of studies their pain wasn't clearly higher. In UKUFF, people with a re-tear still did better than those whose tear couldn't be repaired, and those whose repair healed did best. The AAOS says re-tears often don't cause more pain or loss of function, and another operation is needed only for severe pain or loss of function.

So a re-tear on a scan describes the tendon, not how well you will do. A sudden loss of strength after a fall or a jerk is different: tell your team the same day.

Sources 11, 4, 3

If you had a decompression, not a repair

Shaving bone from under the roof of the shoulder (a subacromial decompression) is a different operation. In a UK trial of 313 people whose cuff wasn't torn through, it did no better than a placebo operation, and recovery from it doesn't need the protections on this page. Our shoulder impingement guide covers it. A tear that can't be repaired is sometimes treated with a reverse shoulder replacement, which has its own guide.

Source 12

Moving early is safe, and adds a little movementStrong evidence

In a 2021 review of 20 trials (1,841 people), starting rehabilitation early gave a little more movement in the first months (about 7 to 8 degrees more lifting the arm forwards at 6 weeks and 3 months, and about 1 degree at a year), no difference in pain or function at most points, and no more re-tears. Most trials were at high or unclear risk of bias, and most of their early programmes moved the arm for people rather than having them use it.

A review of 8 studies (756 people) that tested early active movement found more movement in the first 6 months, slightly worse shoulder-related quality of life at 6 weeks, and no difference in healing; the authors judged that none of the differences mattered in practice. An older review of 12 trials found a slightly higher re-tear rate with early rehabilitation that could have been chance, so a small rise can't be ruled out. A 2025 review of 11 trials found early exercise cut post-operative stiffness by about two-thirds.

Sources 13, 14, 15, 16

Two well-known trials: no difference by 1 to 2 yearsEmerging evidence

In a trial of 206 people (average age 56), weaning off the sling and doing pain-free active movement in the first 6 weeks gave better lifting at 6 weeks than 6 weeks in a sling, and no differences in movement, pain, strength or quality of life after that, up to 2 years. The authors suggest allowing pain-free active movement in the first 6 weeks.

In a trial of 124 people under 65 with tears smaller than 3 cm, early movement with help gave better lifting and turning at 3 months than 6 weeks immobilised, and the same function and strength from 6 months to 2 years; 92% of repairs had healed. Both trials were mostly small and medium tears.

Sources 8, 17

A UK trial is testing early, self-directed recoveryEmerging evidence

In the RaCeR pilot trial (Rehabilitation after Rotator Cuff Repair), 73 adults at 5 NHS hospitals in England were advised either to take the sling off as soon as they could and move as their symptoms allowed, or to wear it for 4 weeks. The early group spent much more time out of the sling, and there were 7 full-thickness re-tears in the early group against 11 in the sling group. As a pilot, it was set up to test whether a full trial could work, not to show which approach is better.

The full RaCeR 2 trial, with 638 people, compares that approach with the NHS standard of a sling for at least 4 weeks; its results aren't out yet. Meanwhile, in a survey of 76 UK shoulder surgeons, 87% agreed there is real uncertainty about the best approach, and in a survey of 129 clinicians, most still advised 4 to 6 weeks in a sling, active movement from 4 to 6 weeks, and resisted exercise and driving from 7 to 12 weeks.

Sources 18, 19, 20, 21

The sling: plans differ, and a pillow brace adds nothingEmerging evidence

NHS hospitals' plans range from 4 weeks in a sling to up to 6 weeks, day and night, and the AAOS gives 4 to 6 weeks. A US therapists' consensus (read as a summary) describes 2 weeks of strict immobilisation, then protected movement with help from weeks 2 to 6, active movement after that, and strengthening from week 12. Timings vary with the size of the tear and your surgeon's protocol, which comes first.

In 4 trials (224 people), an abduction brace, a sling with a pillow holding the arm a little away from the body, gave no better pain, function or healing than a simple sling. Use what your team gives you.

Sources 1, 5, 2, 3, 22, 23

Keep pendulums smallEmerging evidence

In a laboratory study of 13 healthy people, large pendulum swings, about 50 cm across, and even having a drink made the supraspinatus and infraspinatus work at more than 15% of their maximum, and large swings worked the supraspinatus harder than small ones, about 20 cm across. The authors suggest larger swings may load the tendon more than is wise early after a repair, so small, relaxed swings moved by your body are the safer choice.

Source 24

Getting back to work and sportEmerging evidence

Across 13 studies of 1,224 people, about 62% returned to their previous level of work, at about 8 months on average, and fewer of those in physically demanding jobs. Across 25 studies of 859 people, about 85% returned to sport and about 66% to the same level, after 4 to 17 months; only about half of competitive athletes got back to the same level. A 2023 review of 11 studies found return-to-sport rates of 50% to 100%. These are observational studies, mostly from specialist centres.

Sources 25, 26, 27

Typical precautions, which your team may change

These come from NHS hospital leaflets, and your own team's list comes first. Most are about protecting the repair for the first 6 to 12 weeks.

• No lifting with the operated arm, not even a kettle, for 8 to 12 weeks, and nothing heavy, such as digging or manual work, for 4 to 6 months.
• Don't use the arm with the elbow away from your body for the first 3 to 6 weeks: one NHS hospital says this may damage the repair.
• No pushing up from a chair, bed or bath with that arm: until 12 weeks in one plan, with care for 3 months in another.
• Don't lie on the operated side for 6 weeks, and don't let the elbow drop across the front of your body at night.
• Wear the sling at night as well as by day for as long as your team says.

Sources 1, 5, 2

Typical timings, which your team may change

These come from NHS hospital leaflets, and they differ. Your own team's timings come first.

• Sling: from 4 weeks to up to 6 weeks, day and night.
• Driving: never in a sling, which NHS leaflets say is illegal; usually between 6 and 12 weeks, once you can control the car safely. You don't need to tell the Driver and Vehicle Licensing Agency (DVLA) unless your recovery is likely to affect your driving for more than 3 months, but you must stay insured.
• Work: leaflets differ on desk work, from about 3 weeks for work close to the body in one hospital's plan to 6 to 8 weeks in another's. Most people are back within 2 months; heavy or overhead work can take 3 to 6 months.
• Swimming: a gentle modified stroke at 6 to 8 weeks and front crawl from about 12 weeks.
• Golf: from about 6 months. Contact sport: at about 9 to 12 months.

Sources 1, 5, 2, 28

Sleeping

Wear the sling at night for as long as your team says. Lying on your back with a small towel or pillow under the upper arm is often the most comfortable. If you lie on your good side, rest the operated arm on pillows in front of you so the elbow doesn't drop across your body.

Sources 1, 2

The lying assisted arm raise (written steps)

Many NHS leaflets use this as the first arm lift. Do it only when your team includes it, and only to the height they allow. The 3D figure doesn't show it yet; the table slide works a similar range sitting up.

• 1. Lie on your back on a bed with your knees bent and a pillow under your head. Get into position without pushing up through the operated arm.
• 2. Hold the wrist of the operated arm with your good hand. Keeping the operated elbow bent at first makes it easier.
• 3. Let the good arm lift the operated arm slowly towards the ceiling, keeping the operated arm as relaxed as you can.
• 4. Stop at the height your team allows, short of a stretch, and lower slowly.
• 5. 5 times, building up, 3 to 4 times a day: one NHS hospital's dose.

Source 1

Gentle holds against your other hand (written steps)

These keep the shoulder muscles working without moving the joint. One NHS hospital adds them in its second phase, once you're told you can start, and the NHS standard plan in a UK trial uses them at about 6 to 8 weeks. Do them only if your team gives them to you.

• 1. Lie on your back, or sit or stand, with the operated elbow bent to a right angle and tucked against your side.
• 2. Hold the wrist of the operated arm with your good hand.
• 3. Turning in: try to move the operated hand in towards your tummy, but stop it moving with your good hand. Hold 5 seconds.
• 4. Turning out: try to move the operated hand outwards, again stopping it with your good hand. Hold 5 seconds.
• 5. Use gentle effort: 5 of each, building gradually to 20, as one NHS hospital advises. After a repair of the front tendon (subscapularis), leave out the turning-in hold until your team says.

Sources 1, 19