Is it a rotator cuff problem or frozen shoulder?
Both cause shoulder pain and both make it hard to use the arm, but the mechanics are different. Rotator cuff problems are tendon and muscle driven, which means the cuff hurts and weakens but the joint can usually still be moved by someone else. Frozen shoulder (adhesive capsulitis) is a capsule problem, which means the whole joint stiffens up so badly that even a physio cannot move it past a certain point. That is the crucial separator, and it is why they need very different plans.
Written by Kareem Hassanein, Registered Physiotherapist. Burlington, Ontario. This is a guide, not a diagnosis. A brief in-person exam sorts these out reliably.
Side by side
The patterns that separate rotator cuff from frozen shoulder in clinic. Read across each row and compare.
| Aspect | Rotator cuff | Frozen shoulder |
|---|---|---|
| Typical age and onset | Any adult age. Often starts after a specific overuse episode, a lift that went badly, or a fall onto an outstretched arm. | Usually 40 to 60. Often starts for no obvious reason. Strongly linked to diabetes, thyroid disease, and the non-dominant side of a recent upper-limb injury. |
| How the range of motion looks | You can often move the arm to nearly full range when I lift it for you, even if it hurts. Active range is limited mostly by pain and weakness, not a hard stop. | A clear hard block. Both your active and my passive movement stop well short of normal, particularly on external rotation with the arm at your side. |
| External rotation (turning the palm out with the elbow tucked) | Usually preserved unless there is a large cuff tear affecting infraspinatus specifically. | Dramatically lost. This is the single most useful bedside finding. A loss of passive external rotation is the hallmark of frozen shoulder. |
| Night pain | Common, especially lying on that side. Usually reduces when you find a comfortable position. | Classic, severe, and position-independent. Often wakes people multiple times a night during the painful phase, even on the other side. |
| Strength | Weakness on specific tests is central. External rotation, elevation in the plane of the scapula, and lift-off tests pick up cuff deficits. | Strength within the available small range is reasonable. The problem is range, not power. If I bring the arm into a position where strength can be tested, it usually grades close to normal. |
| Timeline | Can settle in weeks to months with the right loading program. A true full-thickness tear may not fully close but is often manageable without surgery. | Follows a slow three-phase course: painful (2 to 9 months), stiff (4 to 12 months), thawing (5 to 24 months). Total natural history is 1 to 3 years, though rehab accelerates recovery. |
| What imaging usually shows | MRI or ultrasound shows tendinopathy or partial or full-thickness tears of supraspinatus, infraspinatus, subscapularis, or teres minor. | Imaging is often unremarkable. The diagnosis is clinical. MRI arthrogram can show a thickened, contracted capsule, but scans are not usually needed to confirm it. |
Typical age and onset
Rotator cuff
Any adult age. Often starts after a specific overuse episode, a lift that went badly, or a fall onto an outstretched arm.
Frozen shoulder
Usually 40 to 60. Often starts for no obvious reason. Strongly linked to diabetes, thyroid disease, and the non-dominant side of a recent upper-limb injury.
How the range of motion looks
Rotator cuff
You can often move the arm to nearly full range when I lift it for you, even if it hurts. Active range is limited mostly by pain and weakness, not a hard stop.
Frozen shoulder
A clear hard block. Both your active and my passive movement stop well short of normal, particularly on external rotation with the arm at your side.
External rotation (turning the palm out with the elbow tucked)
Rotator cuff
Usually preserved unless there is a large cuff tear affecting infraspinatus specifically.
Frozen shoulder
Dramatically lost. This is the single most useful bedside finding. A loss of passive external rotation is the hallmark of frozen shoulder.
Night pain
Rotator cuff
Common, especially lying on that side. Usually reduces when you find a comfortable position.
Frozen shoulder
Classic, severe, and position-independent. Often wakes people multiple times a night during the painful phase, even on the other side.
Strength
Rotator cuff
Weakness on specific tests is central. External rotation, elevation in the plane of the scapula, and lift-off tests pick up cuff deficits.
Frozen shoulder
Strength within the available small range is reasonable. The problem is range, not power. If I bring the arm into a position where strength can be tested, it usually grades close to normal.
Timeline
Rotator cuff
Can settle in weeks to months with the right loading program. A true full-thickness tear may not fully close but is often manageable without surgery.
Frozen shoulder
Follows a slow three-phase course: painful (2 to 9 months), stiff (4 to 12 months), thawing (5 to 24 months). Total natural history is 1 to 3 years, though rehab accelerates recovery.
What imaging usually shows
Rotator cuff
MRI or ultrasound shows tendinopathy or partial or full-thickness tears of supraspinatus, infraspinatus, subscapularis, or teres minor.
Frozen shoulder
Imaging is often unremarkable. The diagnosis is clinical. MRI arthrogram can show a thickened, contracted capsule, but scans are not usually needed to confirm it.
What I check in person to separate them
These are the clinical tests I actually run in the first visit. You cannot do them all on yourself reliably, but understanding what they look for helps explain why an in-person exam sorts these so quickly.
Passive external rotation with the arm at the side
I hold your elbow into your side, support the forearm, and rotate the hand outward. A firm capsular block well short of normal (usually less than 30 degrees on the affected side compared with the other arm) is the strongest bedside marker of frozen shoulder.
Drop arm / empty can / external rotation strength (rotator cuff battery)
These isolate the main cuff muscles. Weakness or pain on resisted external rotation and elevation in the plane of the scapula is classic for rotator cuff pathology.
Hawkins-Kennedy and Neer impingement signs
Positive in rotator cuff and subacromial pain. Can also be positive in the painful phase of frozen shoulder, so they are supportive rather than definitive.
Active versus passive range comparison
In rotator cuff problems, what I can move the arm through passively is usually much greater than what you can move actively. In frozen shoulder, the two are almost identical and both are limited.
Scapular assist and relocation tests
Improving active range by manually positioning the shoulder blade suggests scapular and cuff contributions. A frozen capsule does not budge with scapular assistance.
Which pattern fits you better?
Plain-language routing. This is not a diagnosis, and real patients often sit between the two, but the language below is a reasonable starting point.
Rotator cuff
Your pattern more closely matches a rotator cuff problem if the pain is sharpest with specific movements (reaching overhead, behind the back, or across the body), you can still get reasonable passive range when someone else lifts your arm, external rotation with the elbow at your side is preserved, and specific cuff tests reproduce weakness or pain. A recent fall, lift, or ramp-up in overhead activity often sits in the history.
Read the rotator cuff pageFrozen shoulder
Your pattern more closely matches frozen shoulder if the shoulder has progressively stiffened up over weeks to months with no single clear injury, night pain is severe, and both active and passive external rotation at the side are strikingly reduced. A history of diabetes or thyroid disease raises the probability. The arm may actually hurt less as it stiffens, which misleads some people into thinking they are getting better.
Read the frozen shoulder pageIf you still cannot tell
The two can look similar in the early painful phase, which is exactly when an honest assessment helps. I check passive external rotation first (it does most of the separating work), then run the cuff battery, compare active and passive range, and ask targeted history questions. If I am still unsure, I will tell you that and re-check at the next visit, because frozen shoulder declares itself more clearly with time.
When both are going on
A stiff, painful shoulder after a rotator cuff strain can progress into secondary stiffness that looks like early frozen shoulder. It also works the other way: people with true frozen shoulder often have co-existing cuff tendinopathy from years of compensation. Treating one and ignoring the other is a common reason shoulders do not fully recover.
Questions patients ask about telling these apart
My shoulder is stiff and painful. Is it just a rotator cuff injury that needs more time?
Maybe, but there is a specific test that helps. Try to externally rotate your shoulder with the elbow tucked into your side, then compare it to the other arm. If that motion is strikingly limited and painful on the affected side and the other side looks normal, frozen shoulder is very much in the picture. If external rotation at the side is similar to the other arm, rotator cuff issues become more likely. Either way, an exam sorts it in one visit.
Does frozen shoulder get better on its own?
Most cases eventually resolve, but the natural timeline is long. Published series show one to three years from onset to full recovery without treatment, with meaningful functional loss in the stiff phase. Structured physiotherapy focused on capsular mobility, pain control, and progressive loading consistently shortens this and reduces residual stiffness.
Should I get an MRI before starting physiotherapy?
For most shoulder pain, no. Frozen shoulder is a clinical diagnosis. Rotator cuff problems are often managed well without imaging, and incidental findings on MRI in pain-free adults are extremely common. I order imaging when the exam raises specific concerns, when symptoms are not improving on the expected timeline, or when surgery is a real consideration.
Can I strengthen my way out of frozen shoulder?
Not in the usual strengthening sense. In the painful phase, loading into an already irritable, contracted capsule tends to flare things. The early work is about restoring range through careful mobilisation, sleep-tolerant positions, and gentle movement. Strengthening comes in later as range returns. The order matters, which is the opposite of most rotator cuff rehab.
What if I have both a rotator cuff tear and a frozen shoulder?
Common, and manageable. The plan addresses the stiffer pattern first because you cannot strengthen into range you do not have. As capsular mobility improves, the rotator cuff rehab layers in. Skipping the capsular work early usually stalls the whole rehab.
I have diabetes. Does that change anything?
Yes. Frozen shoulder is several times more common in people with diabetes, it tends to be more severe, and it takes longer to recover. That is not a reason to abandon rehab, it is a reason to start earlier, pace more carefully, and protect sleep aggressively through the painful phase.
Evidence this page draws on
Sources I lean on when separating these two conditions in clinic.
Loss of passive external rotation with the arm at the side is the most useful clinical finding for identifying adhesive capsulitis.
Kelley MJ et al., "Shoulder Pain and Mobility Deficits: Adhesive Capsulitis" (JOSPT Clinical Practice Guideline), 2013.
Adhesive capsulitis progresses through painful, stiff, and thawing phases lasting a combined one to three years without treatment, with residual symptoms possible long-term.
Hand C, Clipsham K, Rees JL, Carr AJ. "Long-term outcome of frozen shoulder." Journal of Shoulder and Elbow Surgery 2008; 17(2): 231-236.
Rotator cuff tendinopathy and tears are often managed successfully with structured exercise and loading, with surgical and non-surgical outcomes comparable for many degenerative tears at two to five years.
Kuhn JE et al., "Effectiveness of physical therapy in treating atraumatic full-thickness rotator cuff tears." Journal of Shoulder and Elbow Surgery 2013; 22(10): 1371-1379.
Treatments that commonly sit inside either plan
The specific mix depends on the assessment and your goals. These are the pieces I draw from most often for both conditions.
Exercise Therapy
Personalized exercise programs designed to restore strength, flexibility, and function.
Joint Mobilization
Graded techniques to restore joint movement and reduce stiffness.
Dry Needling
Precise needle therapy targeting trigger points for effective pain relief and improved muscle function.
Soft Tissue & Myofascial Therapy
Targeted hands-on techniques to address muscle tension, pain, and movement restrictions.
Post-Surgical Rehabilitation
Evidence-based recovery programs following surgery to restore function and strength.
Book an assessment
If this page has helped you narrow things down, or if it has left you wanting a proper exam, I see patients at Endorphins Health & Wellness Centre in Burlington. Direct insurance billing is available, and a physician referral is not required.
4631 Palladium Way, Unit 6
Burlington, ON L7M 0W9
Direct billing. No referral needed.
- Monday1:30 PM - 7:30 PM
- Tuesday3:30 PM - 7:30 PM
- Wednesday2:00 PM - 7:30 PM
- Thursday1:30 PM - 7:30 PM
- Friday2:00 PM - 7:30 PM