Piriformis Muscle: Anatomy, Function and What the Research Actually Shows

Key Takeaways

In a dissection study of 124 lower limbs, the piriformis muscle was present in 96.8%. The classic origin on the anterior sacrum between S2 and S4 was the most common pattern, but it was found in only 52 of the 120 limbs in which the muscle was present (43.3%).[1]
A meta-analysis of 6,062 cadavers put the prevalence of piriformis and sciatic nerve anomalies at 16.9%, a figure not significantly different from the 16.2% seen in surgical case series of piriformis syndrome.[2]
Biomechanical models commonly describe the piriformis as switching from an external to an internal rotator as the hip flexes. A 2026 cadaver study concluded that its findings, considered alongside the documented variability of the tendon attachment, call such simplified models into question.[3]
At 90° of hip flexion the muscle lay directly behind the hip joint in cadavers, and cadaver work suggests it may help restrict internal rotation at 0° and 60° of hip flexion, although the findings differed depending on the order in which the muscles were cut.[4][5]
In 21 healthy young men, shear wave elastography showed the highest piriformis muscle shear elastic modulus, which the authors interpreted as the greatest stretch, in maximal external rotation combined with 110° of hip flexion and 40° of hip adduction.[6]
In a cadaver study, 95% of ultrasound-guided piriformis injections were judged accurate, compared with 30% of fluoroscopically guided, contrast-controlled injections.[7]
In my clinical experience, deep buttock pain is a common reason for consultation, but its exact source often remains unclear, and clear radiating pain has surprisingly often turned out to be caused by a lumbar disc herniation.
In my practice, active physiotherapy, including isometric exercise, often seems to help more in the long term than passive treatments, which may bring quick but short-lived relief.
In the patients I see, static working postures are a common background, including home-office work at a kitchen table.

Introduction: Piriformis Muscle

The piriformis muscle is often discussed as a possible source of buttock pain and sciatica-type symptoms, and piriformis stretches are a familiar part of many rehabilitation routines. Yet the muscle itself is usually described in simplified terms: a pear-shaped muscle running from the sacrum to the greater trochanter, and an external rotator that reverses its action when the hip bends.

The piriformis is a small muscle, but it is often the subject of considerable clinical deliberation. Pain in the buttock region, often described as deep gluteal pain, is a common reason for consultation, and in my clinical experience the piriformis muscle comes up in many of these conversations.

Anatomical research paints a more complicated picture. Dissection studies show that the origin, insertion and relationship to the sciatic nerve vary considerably between individuals, and some widely repeated functional claims are less settled than often assumed.

This article covers the anatomy of the piriformis muscle, how variable it is, what it does, what electromyography (EMG) and stretching studies show, and why its deep location matters when the muscle is injected. The clinical condition attributed to it is covered separately in our article on piriformis syndrome.

Piriformis Muscle Anatomy: Origin, Insertion and Nerve Supply

In the standard description, the piriformis muscle arises from the anterior part of the sacrum and leaves the pelvis through the greater sciatic foramen. As it passes laterally towards the tip of the greater trochanter, it narrows into a tendon that lies on the posterosuperior part of the hip joint capsule. It is supplied by the nerve to piriformis, which comes from the ventral rami of S1 and S2 in the sacral plexus.[1]

The insertion has been studied in detail in relation to hip replacement surgery. In a dissection of 15 cadavers, the tendon attached to the crest of the greater trochanter, above the trochanteric fossa. Its footprint covered the summit and medial aspect of the greater trochanter, with a variable additional attachment to the fibrous hip capsule.[4] The same authors pointed out that the “piriform fossa”, a term orthopaedic surgeons use for the trochanteric fossa, normally has nothing to do with where the piriformis tendon attaches, and concluded that the muscle’s anatomy is often described inaccurately in the surgical literature.[4]

An earlier study of 112 cadaveric specimens found four insertion patterns. In 53.57%, a rounded piriformis tendon reached the upper border of the greater trochanter on its own. In 29.46%, it first joined the superior gemellus tendon and then the obturator internus tendon before inserting on the medial surface of the greater trochanter. In 13.39%, the tendon fused with both the obturator internus and gluteus medius tendons, and in 3.57% it fused with the gluteus medius alone.[8]

From a clinical perspective, the location of buttock pain gives only limited information about its source. In my clinical experience, many patients point to pain around the gluteus medius insertions, while a smaller group describes pain deeper in the buttock, roughly where the piriformis muscle lies anatomically. In these clinical situations, it often remains unclear whether the pain actually originates from the piriformis or from another muscle in the same region, even when the symptoms clearly localise to this area.

How Variable Is the Piriformis?

What its authors describe as the first comprehensive classification comes from a 2024 dissection of 124 lower limbs from 62 formalin-fixed cadavers from a Central European population. The piriformis muscle was present in 120 limbs (96.8%) and absent in 4, with two cadavers lacking it on both sides.[1]

The proximal attachment varied in four ways:[1]

Anterior surface of the sacrum, between S2 and S4: 43.3%
Sacrotuberous ligament: 33.3%
Gluteal surface of the ilium near the margin of the greater sciatic notch: 13.3%
Ilium near the greater sciatic notch together with the gluteus medius: 10%

Distally, a single tendon was found in 87.5% of limbs, a two-part tendon in 9.2% and a three-part tendon in 3.3%. The piriformis was fused with neighbouring muscles (gluteus medius, superior gemellus, or superior gemellus and gluteus minimus) in 31.7% of limbs, and a completely double piriformis was found in 5 limbs.[1]

The sciatic nerve relationship

In the same series, the sciatic nerve ran below the muscle in 85% of limbs. In 8.3%, the common fibular nerve exited above the muscle and the tibial nerve below it; in 4.2%, the whole sciatic nerve passed above the muscle; and in 2.5%, the common fibular nerve passed between the two bellies of a double piriformis. The authors did not observe the nerve piercing the muscle in any limb.[1]

Pooled data suggest that these figures differ between populations. A 2020 meta-analysis of 44 cadaveric studies found a pooled prevalence of 90% for the typical pattern, an undivided nerve passing below the muscle. Variants were more common among East Asians, at a pooled prevalence of 31%, and the authors reported that in more than 10% of the population the nerve coursed through or above the muscle. These are pooled estimates from cadaveric studies and should not be read as an individual person’s probability.[9]

Whether these variants actually cause symptoms is a different question. A meta-analysis pooling 18 studies and 6,062 cadavers estimated the prevalence of piriformis and sciatic nerve anomalies at 16.9% (95% CI 16.0–17.9%). In surgical case series of patients with piriformis syndrome, the prevalence was 16.2% (95% CI 10.7–23.5%), and the difference was not significant (P = 0.824). The author concluded that the anomaly may be less important in causing piriformis syndrome than previously thought.[2] The 112-specimen study reached a similar conclusion: anatomical causes of piriformis syndrome appeared rare, and the authors called for a more precise work-up to exclude more common diagnoses.[8]

What Does the Piriformis Muscle Do?

The piriformis muscle belongs to the lateral rotators of the hip, together with the quadratus femoris, the superior and inferior gemelli, and the obturator internus and externus. In the standard functional description, it rotates the femur laterally when the hip is extended and abducts it when the hip is flexed.[1]

The “reversal of action” debate

A common description is that the piriformis turns from an external rotator into an internal rotator as hip flexion increases. Biomechanical models commonly describe it this way, but they often assume that the distal tendon attaches in the same place in everyone.[3]

A 2026 study tested part of this assumption in 63 femora from 38 embalmed donors by relating the position of the piriformis tendon on the greater trochanter to femoral torsion. Across the whole sample, there was no significant relationship (r = 0.13, p = 0.31). Left femora showed a moderate positive correlation (r = 0.38, p = 0.04), while right femora showed no meaningful association. The authors concluded that, together with the documented variability of the distal attachment, these findings call into question simplified models that assume a predictable reversal of piriformis function with hip flexion.[3] The study does not show that reversal never occurs; its conclusion is that piriformis biomechanics should be interpreted more cautiously. It is also worth noting that the study measured tendon position and femoral torsion; it did not measure the piriformis muscle’s rotational action directly.

A possible role in hip stability

Cadaver work also points to a stabilising role. When the hip was flexed to 90°, the piriformis lay directly behind the hip joint, and the authors suggested it may plausibly contribute to joint stability.[4] In a separate study of nine cadaveric hips, the external rotators were cut one by one while an internal rotation force was applied at 0°, 30° and 60° of hip flexion. The ischiofemoral ligament was the main restraint against internal rotation, and the authors concluded that the piriformis and obturator internus may restrict internal rotation at 0° and 60° of flexion. However, the pattern of statistically significant findings differed between the two cutting orders, and this small cadaveric sample does not establish the piriformis muscle’s independent contribution.[5]

What EMG shows

In a study using fine-wire electrodes inserted directly into the piriformis of 11 healthy men, activity was measured during seven prone and side-lying hip movements. Piriformis activity was highest during prone hip extension performed in external rotation, a movement that also strongly activated both the upper and lower portions of the gluteus maximus.[10] With only 11 healthy male participants, the study describes muscle activation patterns; it does not show what training these movements would achieve or how they affect symptoms.

Stretching the Piriformis Muscle: What Positioning Studies Show

Static stretching of the piriformis muscle is one of the physical therapy treatments used for piriformis syndrome, yet the most effective stretching position had not been clearly established in living people.[6] Two small studies have approached this question using different measures.

In 21 healthy young men (mean age 22.7 years), shear wave elastography was used to measure the shear elastic modulus of the piriformis in 12 hip positions, as a way of identifying where the muscle is stretched most. The modulus was significantly higher at 40° than at 20° of hip adduction, and higher at 20° than at 0°, and it was higher in external rotation than in internal rotation. The highest shear elastic modulus was found in maximal external rotation combined with 110° of hip flexion and 40° of hip adduction, which the authors interpreted as the position in which the muscle was most stretched.[6]

An earlier study used CT-based three-dimensional modelling in seven subjects to estimate piriformis elongation during two conventional stretches, which combined hip flexion, adduction and external rotation in different orders. Both lengthened the muscle by approximately 12%, and the anteversion of the femoral head and greater trochanter influenced how much the muscle lengthened. A computer model then identified positions with deeper hip flexion, for example 115° of flexion, 40° of external rotation and 25° of adduction, that increased piriformis elongation by 30–40% compared with the conventional stretches.[11]

Both studies were small and measured muscle length or stiffness rather than symptoms. They identify positions associated with greater stiffness during stretch or greater modelled lengthening, not whether these positions change pain or function. For the wider evidence on what stretching does and does not achieve, see our article on stretching for athletes.

Injections: Why Precise Location Matters

The piriformis lies beneath the gluteus maximus, close to the sciatic nerve. In a cadaver study, 20 piriformis muscles in 10 unembalmed cadavers were injected with liquid latex using both ultrasound guidance and fluoroscopic guidance with contrast control. Ultrasound-guided injections were judged accurate in 19 of 20 cases (95%), compared with 6 of 20 (30%) for the fluoroscopic technique (P = .001). In 13 of the 14 missed fluoroscopic injections, and in the single missed ultrasound injection, the latex ended up in the gluteus maximus. The remaining missed fluoroscopic injection was found within the sciatic nerve.[7]

This study assessed accuracy from where the injected latex was found on dissection, not clinical treatment outcomes, and all injections were performed by one experienced operator. It does show that, in this model, injections aimed at the piriformis could end up in a neighbouring muscle, particularly with fluoroscopic guidance, which may matter when interpreting both the diagnostic response to an injection and treatment studies. The clinical evidence on injection outcomes is discussed in our piriformis syndrome article.

In practice, some patients ask for corticosteroid injections, and I do sometimes inject the painful area. In my practice, corticosteroid injections are not a first-line option for this kind of pain, but in my clinical experience some patients report benefit from them. I cannot say how much of that effect reflects placebo and how much a genuine treatment effect.

Anatomy Is Not a Diagnosis: The Link to Piriformis Syndrome

Piriformis syndrome is described as a form of sciatica caused by compression of the sciatic nerve by the piriformis muscle. A 2019 narrative review called it a relatively uncommon but not insignificant cause of sciatica, and noted that diagnosis is complicated by the wide differential of low back and buttock pain, in which many conditions share overlapping symptoms.[12] The authors of the 124-limb dissection study likewise cautioned against overdiagnosis, listing lumbar canal stenosis, disc inflammation and pelvic causes among the conditions that can mimic it.[1]

Many patients also arrive with an explanation already in mind. They may have read or heard, or been told by a chiropractor, that they have sacroiliac joint dysfunction or that the joint is somehow out of place, and some have already sought chiropractic treatment for it. The pain is indeed often felt in the same area where the SI joint is typically located. In my clinical experience, however, patients rarely get significant long-term help from chiropractic manipulation, and in my view the idea that the SI joint becomes subluxated in a way that chiropractic treatment could correct is not well supported, although this article does not review that evidence. Beliefs about this region are nevertheless common and can be difficult to change.

A second group of patients has radiating, radicular pain. The sciatic nerve does run close to the piriformis muscle, and its course varies between individuals, as the dissection studies above show. In my clinical experience, however, surprisingly often the eventual explanation for clear radicular symptoms has been a lumbar disc herniation. Radicular pain caused purely by the piriformis is, in my view, very difficult to confirm clinically, so when a patient has clear radiating pain, I have usually referred them for an orthopaedic assessment at least once. Deep gluteal syndrome does appear to exist as a clinical entity, but in my understanding it is not particularly well recognised in Finnish medicine.

In other words, knowing the anatomy of the piriformis muscle does not by itself explain a patient’s pain. Other common explanations for pain around the buttock and hip are discussed in our articles on whether the SI joint can be “out of place” and on gluteal tendinopathy.

Deep Buttock Pain in Clinical Practice

Many patients with buttock-region pain sit for much of their working day, although the area can also become painful in people who stand at work. In my clinical experience, the main problem is often that working postures are static, with too little variation. Most of my patients now have height-adjustable desks at work, but problems often arise when they work from home, where the workstation may be a kitchen table and a kitchen chair rather than an ergonomically planned setup. In practice, an occupational health physiotherapist can help by reviewing working postures, posture changes and ergonomics, so that the load becomes less static and the patient copes better with the working day. These are usually the kinds of changes my patients benefit from.

I refer most of these patients to physiotherapy. In my clinical experience, they appear to benefit most from active physiotherapy, including isometric exercises. Pain medication and muscle relaxants are sometimes prescribed, but I explain to patients that they are not a final solution. Passive physiotherapy is not ruled out, and many patients find it helpful, but in my clinical experience it tends to work best as an addition to active physiotherapy rather than as a replacement for it.

The situation is complicated by the fact that massage, sometimes stretching, and chiropractic treatment often bring immediate relief, which in my experience does not last long. Active exercise, by contrast, can make the muscles sore at first, which may discourage patients from continuing. This can steer patients back towards passive treatments and medication, which in the long term may not be the best option for them.

Patients often ask whether they are allowed to exercise. In my clinical experience, staying active often appears worthwhile, at least in the long term. Some patients notice that activity increases their pain, but in my experience this is usually only at first, whereas staying still in a static position tends to make things worse. Many patients also tell me that exercise seems to help their symptoms, while the pain often worsens as soon as they stop moving or sit down.

Conclusion: Piriformis Muscle

The piriformis muscle is far less uniform than its standard description suggests. In dissection studies, its origin, insertion and fusions with neighbouring muscles vary widely, the muscle is occasionally absent or doubled, and a pooled analysis found piriformis and sciatic nerve anomalies in 16.9% of cadavers. That proportion was not significantly different in surgical case series of piriformis syndrome, which suggests anatomical variants may be less important in causing the condition than once thought.

Functionally, the piriformis is a lateral rotator and abductor that sits directly behind the flexed hip and may help restrict internal rotation. The popular “reversal of action” model is less settled than often assumed. In small positioning studies, combinations of deep hip flexion, adduction and external rotation produced the greatest stiffness on elastography or the greatest modelled lengthening, while a cadaver injection study showed that injections aimed at this deep muscle can miss it. For anyone dealing with buttock pain, the anatomy is a useful starting point, but it is not a diagnosis.

In my clinical experience, many patients with deep buttock pain sit or stand in static positions for much of the working day, and it is often difficult to say which structure is producing the pain. Clear radiating symptoms have, in my practice, surprisingly often led back to the lumbar spine. Active rehabilitation has, in my experience, often helped more in the long term than passive treatments alone, even though it can initially increase soreness.

Piriformis Muscle: Anatomy, Function and What the Research Actually Shows

Table of Contents

Key Takeaways

Introduction: Piriformis Muscle

Piriformis Muscle Anatomy: Origin, Insertion and Nerve Supply