The crank test is a hands-on clinical examination used to detect tears in the glenoid labrum, the ring of cartilage that lines the shoulder socket. It gained attention in the mid-1990s after its developers reported striking accuracy, but the picture has grown murkier since then. Depending on which study you read, the test’s ability to correctly identify a labral tear ranges from under 50% to above 90%, a spread wide enough to fuel genuine debate about how much weight a clinician should give the result.
What the Crank Test Actually Involves
During a crank test, you typically stand or lie on your back with your arm raised to about shoulder height and your elbow bent at a right angle. The examiner stabilizes your shoulder blade with one hand, then uses the other hand to push along the length of your upper arm, compressing the humeral head into the socket. While maintaining that pressure, the examiner rotates your arm fully inward and outward. The test is considered positive if you feel pain during the rotation, especially if it reproduces the deep, clunking, or catching sensation you have been experiencing. Some clinicians also note whether they feel a mechanical click through their hands.
The test’s logic is straightforward: by pressing the ball of the shoulder firmly into the socket and then spinning it, a torn labral edge gets pinched or ground against the joint surfaces, provoking symptoms. That compression-plus-rotation combination is what distinguishes the crank test from other shoulder maneuvers that stress the labrum through different loading directions.
The Original Study and Its Eye-Catching Numbers
The crank test was introduced in a 1996 prospective study that evaluated patients who went on to have arthroscopic surgery, making it possible to compare the test’s prediction to what was actually found inside the joint. In that initial report, the sensitivity was 91%, the specificity was 93%, the positive predictive value was 94%, and the negative predictive value was 90%. The authors concluded that the crank test met the criteria for a single examination maneuver that is “highly accurate for the preoperative diagnosis of glenoid labral tears.”1PubMed. A prospective evaluation of a new physical examination in predicting glenoid labral tears
Those numbers were impressive enough to earn the test a place in shoulder examination protocols worldwide. If a single bedside maneuver could catch nine out of ten labral tears and avoid false alarms 93% of the time, it would rival the performance of imaging studies costing far more time and money.
Why Later Studies Painted a Different Picture
The optimism faded when independent groups tried to replicate those results. A study comparing the crank test and the O’Brien active compression test against arthroscopic findings in 65 patients found the crank test was positive in only about 45% of them. Its sensitivity dropped to 46%, specificity to 56%, and positive predictive value to just 41%.2PubMed. The crank test, the O’Brien test, and routine magnetic resonance imaging scans in the diagnosis of labral tears That is essentially a coin flip for both catching tears and ruling them out.
A systematic review examining clinical tests for superior labral anterior-posterior (SLAP) lesions explicitly noted the gap: out of the studies reviewed, only the original group of authors reported the high diagnostic accuracy values of 91% sensitivity and 93% specificity. Other investigators, working with different patient populations and examiners, could not match those figures.3PubMed. Diagnostic accuracy of clinical tests for superior labral anterior posterior lesions: a systematic review A separate systematic review echoed this finding, identifying “varying results of accuracy in the Crank test” when it was evaluated in more than one study, while acknowledging that the original report’s positive likelihood ratio of roughly 13.6 was among the highest of any shoulder special test.4PubMed. The validity and accuracy of clinical tests used to detect labral pathology of the shoulder–a systematic review
So what happened? There is no single smoking gun, but several factors likely explain the discrepancy. The original study’s patient population was drawn from a surgical practice where the prevalence of labral tears was high, and the examiners had developed the test themselves, meaning they had extensive practice performing it in exactly the way they intended. When a test migrates to other clinics, different examiners, and different mixes of shoulder pathology, its accuracy often drifts downward. This is a well-known pattern across orthopedic special tests, not something unique to the crank test.
Agreement Between Examiners
Even before worrying about whether the test catches real tears, there is the question of whether two different examiners performing the crank test on the same patient will agree on whether it is positive. A study of inter-examiner reliability for common shoulder special tests found that the crank test achieved only “fair” agreement, with a kappa value of 0.36 to 0.38.5PubMed. Interexaminer reliability of orthopaedic special tests used in the assessment of shoulder pain For context, a kappa of 1.0 would mean perfect agreement, and anything below about 0.40 is generally considered only fair. The active compression test landed in the same range.
This is a practical problem. If one physical therapist calls your crank test positive and a second calls it negative, the test’s theoretical accuracy becomes academic. Subtle differences in how much axial load is applied, how aggressively the arm is rotated, and how the examiner interprets your pain response all contribute to this variability. It reinforces why most clinicians treat the crank test as one data point in a broader assessment rather than a standalone verdict.
How the Crank Test Stacks Up Against Other Shoulder Tests
The crank test is not the only maneuver aimed at the labrum. The O’Brien active compression test, the anterior slide test, the Speed test, the biceps load tests, and the Jobe relocation test all target overlapping anatomy with different arm positions and loading strategies. A meta-analysis pooling data from six eligible studies found no significant difference in overall accuracy among the active compression test, the crank test, and the Speed test. The anterior slide test, however, performed significantly worse than all three.6The American Journal of Sports Medicine. A Meta-analysis Examining Clinical Test Utility for Assessing Superior Labral Anterior Posterior Lesions
A separate clinical study of seven shoulder tests found that the O’Brien test (about 63% sensitive, 73% specific) and the Jobe relocation test (about 44% sensitive, 87% specific) were statistically correlated with labral tears, while other tests in the battery, including the crank test in that particular sample, did not reach statistical significance for detecting tears.7PubMed. Clinical testing for tears of the glenoid labrum The picture that emerges is one where no single test dominates. Each has studies in which it looks strong and studies in which it looks unremarkable, and their relative performance shifts with the patient population and the examiner.
The more productive question is not “which test is best” but “what happens when you combine them.” Research on combining the crank test or the anterior slide test with a patient’s history of popping or catching in the shoulder found that the combination dramatically boosted specificity. When a patient reported mechanical symptoms and also had a positive crank test, specificity climbed to 0.91; the same combination with a positive anterior slide test pushed specificity to 1.00, meaning essentially no false positives in that sample.8PubMed. Reliability and diagnostic accuracy of history and physical examination for diagnosing glenoid labral tears This aligns with how experienced clinicians actually work: they do not hang the diagnosis on a single maneuver but layer the physical findings together with the patient’s story.
What a Positive or Negative Crank Test Means for You
If your clinician performs a crank test and tells you it was positive, it is reasonable to treat that as one piece of a puzzle rather than a definitive diagnosis. Given the variability in reported accuracy, a positive crank test raises the probability that you have a labral tear, but it does not confirm one. Conversely, a negative result does not confidently rule a tear out, especially if your symptoms otherwise point toward labral pathology.
In practice, a positive crank test in a patient whose shoulder clicks, catches, or gives way during overhead activity will often prompt further workup. That usually means an MRI, sometimes with a contrast injection into the joint (an MR arthrogram), which is considerably better at visualizing labral tissue. The crank test’s real value is in triaging: helping the clinician decide who warrants imaging or a referral and who can be managed conservatively first.
A negative crank test in someone with a vague shoulder ache and no mechanical symptoms can provide some reassurance, though it is hardly ironclad. And if your symptoms strongly suggest a labral tear despite a negative exam, most clinicians will still pursue imaging rather than let the physical test override the clinical picture.
Why Accuracy Numbers Vary So Much Across Studies
The wide range of reported sensitivity and specificity for the crank test is not unusual in orthopedic physical examination research. Several study-design factors drive the spread, and understanding them helps you interpret the numbers without being misled.
Patient selection matters enormously. In a study where every participant is already scheduled for arthroscopy, the prevalence of labral tears is high, often 70% or more. High prevalence inflates the positive predictive value of any test. In a general sports medicine clinic, where many patients have rotator cuff tendinopathy or impingement rather than labral tears, the same test will produce more false positives and look worse on paper. Neither setting is “wrong,” but you cannot compare their numbers directly.
Examiner experience is another variable. The original crank test developers had refined their technique over years before publishing. A resident performing the test for the first time will apply different amounts of force, rotate the arm at different speeds, and interpret pain responses differently. The fair inter-examiner reliability scores confirm that this is not just theoretical. Two examiners doing their best on the same shoulder will disagree roughly a third of the time.
The definition of a positive test also varies. Some protocols count any pain as positive. Others require the reproduction of the patient’s specific symptoms, or insist on a palpable click. Tighter definitions tend to improve specificity (fewer false positives) at the cost of sensitivity (more missed tears), while looser definitions do the opposite. Studies rarely use identical criteria, which makes apples-to-apples comparison difficult.
SLAP Tears Versus General Labral Tears
One common point of confusion is whether the crank test targets the labrum in general or a specific type of labral injury. The test was originally described for glenoid labral tears broadly, but much of the subsequent research focused specifically on SLAP lesions, which involve the superior (top) portion of the labrum where the biceps tendon attaches. This matters because a SLAP tear and a tear along the anterior-inferior labrum (a Bankart-type lesion, often associated with shoulder dislocations) are different injuries with different mechanics.
The meta-analysis comparing special tests specifically for SLAP lesions found that the crank test was in the same accuracy tier as the active compression and Speed tests for that subset of pathology.9The American Journal of Sports Medicine. A Meta-analysis Examining Clinical Test Utility for Assessing Superior Labral Anterior Posterior Lesions But when you broaden the question to all labral tears, some studies found that none of the common tests, individually or in combination, were statistically valid for pinpointing the specific type of tear.10PubMed. Clinical testing for tears of the glenoid labrum In other words, the crank test might flag that something is wrong with the labrum, but it cannot tell your clinician exactly where or how the labrum is torn. That distinction matters for surgical planning, which is another reason imaging typically follows a suspicious exam.
Labral Testing Beyond the Shoulder
The shoulder is not the only joint with a labrum. The hip socket has an analogous cartilage ring, and hip labral tears have become an increasingly recognized source of groin and deep hip pain, particularly in younger active adults. Although the term “crank test” in clinical conversation almost always refers to the shoulder, the concept of compressing a ball-and-socket joint while rotating it to provoke a torn labrum has parallels in the hip examination.
A study evaluating the accuracy of clinical examination for hip labral tears found that physical therapists, orthopedic surgeons, and orthopedic residents all achieved roughly 80 to 85% correct diagnostic accuracy when their clinical impressions were compared to arthroscopic findings.11PubMed Central. Acetabular labral tears: diagnostic accuracy of clinical examination by a physical therapist, orthopaedic surgeon, and orthopaedic residents The hip examination maneuvers are different from the shoulder crank test, typically involving flexion, adduction, and internal rotation of the hip, but the underlying principle is the same: load the joint, rotate it, and see if the torn labrum gets caught. Interestingly, in that study no significant difference in accuracy was found between clinician types, suggesting that experience with the specific maneuvers mattered more than specialty credentials.
When the Crank Test Is Most and Least Useful
The evidence points to a few scenarios where the crank test earns its keep and a few where it adds little. It tends to be most helpful in overhead athletes, such as baseball pitchers, swimmers, and volleyball players, whose repetitive arm motions put the superior labrum under high stress. In this population, the pre-test probability of a labral tear is already elevated, so a positive crank test genuinely shifts the likelihood enough to warrant further investigation. Combined with a history of clicking or catching during the overhead motion, the test’s specificity climbs considerably, as the combined-history research demonstrated.12PubMed. Reliability and diagnostic accuracy of history and physical examination for diagnosing glenoid labral tears
Where it adds less is in the undifferentiated shoulder pain patient with no mechanical symptoms, no history of instability or dislocation, and multiple possible diagnoses on the table. Rotator cuff disease, biceps tendinopathy, and acromioclavicular joint problems can all produce pain during a crank test, leading to false positives that muddy the clinical picture. In these cases, a cluster of tests, a thorough history, and often imaging will do more than any one provocative maneuver.
Patients who have had previous shoulder surgery present another tricky scenario. Scar tissue and altered anatomy can change the way the joint responds to compression and rotation, potentially producing pain or clicking that has nothing to do with a new labral tear. Clinicians generally interpret special tests with extra caution in post-surgical shoulders and lean more heavily on imaging when recurrent pathology is suspected.
The Broader Lesson About Special Tests
The crank test’s story mirrors a pattern seen across musculoskeletal medicine. A new physical examination maneuver is introduced with strong numbers from the group that developed it, other investigators try to replicate those results and get something more modest, and the field eventually settles on using the test as part of a cluster rather than in isolation. This has happened with the McMurray test for meniscal tears in the knee, the Neer and Hawkins tests for shoulder impingement, and the straight-leg raise for lumbar disc herniation. The initial excitement almost always exceeds what independent validation supports.
This does not mean these tests are useless. It means they work best as probability shifters rather than binary verdicts. A skilled clinician who knows the patient’s history, observes how the shoulder moves, and runs several provocative tests is assembling a probabilistic case, much like a detective collecting circumstantial evidence. No single clue closes the case, but together they can point strongly enough in one direction to guide the next step, whether that is imaging, conservative treatment, or a surgical referral. The crank test, despite its uneven track record, remains a standard part of that toolkit precisely because it is quick to perform, costs nothing, and adds information that, in the right context, genuinely changes the clinical calculus.