What Muscle Is on the Inside of Your Knee?

Several muscles converge on the inner side of your knee, not just one. The most visible is the vastus medialis obliquus, or VMO, the teardrop-shaped bulge of muscle just above and inside the kneecap. But beneath the surface, three more tendons merge into a structure called the pes anserinus on the shinbone, and deeper still, the semimembranosus fans out into multiple arms that anchor the back-inside corner of the joint. Which one matters most depends on whether you are asking about appearance, pain, stability, or surgical reconstruction.

The Vastus Medialis Obliquus

The muscle most people notice when they look at the inside of their knee is the lower portion of the vastus medialis, a part of the quadriceps group that runs down the front of the thigh. This lower portion is called the vastus medialis obliquus, often shortened to VMO. Its fibers run at a steep angle toward the kneecap, which distinguishes it from the upper portion of the same muscle (the vastus medialis longus), whose fibers run more vertically. In cadaver studies, roughly nine out of ten limbs show a clear change in fiber direction where the upper portion transitions to the VMO, and many specimens have a distinct fascial plane separating the two portions, suggesting they function almost like two different muscles sharing the same name.¹

The VMO attaches to the inner edge of the kneecap through fleshy fibers and has a small region of direct continuity with the patellar tendon below. Its primary job is pulling the kneecap inward during the last degrees of leg extension, counteracting the outward pull of the vastus lateralis on the opposite side. The oblique fiber angle gives the VMO better leverage for this sideways tug than for generating raw extension force. The upper portion, with its longer fibers and larger cross-sectional area, contributes more to straightening the knee, while the VMO is better suited to stabilizing the kneecap’s position.² Even the fiber-type composition differs between the two portions: the upper segment has a higher proportion of slow-twitch fibers, consistent with a postural endurance role, while the VMO carries more fast-twitch fibers, fitting its job of making rapid corrections to patellar tracking.³

The Pes Anserinus Group

Lower and slightly behind the VMO, three tendons converge on the inner surface of the shinbone in a formation called the pes anserinus, a Latin term meaning “goose’s foot” because of its fan-like shape. These tendons belong to the sartorius, gracilis, and semitendinosus muscles. Each originates from a different place: the sartorius from the front of the pelvis, the gracilis from the pubic bone, and the semitendinosus from the sitting bone at the back of the pelvis. Despite their different starting points, they all end up inserting on the same small patch of tibia just below the inner knee joint line.⁴

Their arrangement is consistent from person to person. From front to back and top to bottom, the order is sartorius first, then gracilis, then semitendinosus. The gracilis and semitendinosus tendons fuse together a few centimeters before they reach the bone and are firmly stuck to the overlying sartorius layer.⁵ Cadaver dissections confirm that the gracilis tendon inserts into the deep fascia surrounding the sartorius, with the semitendinosus accompanying the pair from behind.⁶

Because these three muscles cross both the hip and the knee, they do double duty. At the knee, they flex the joint and rotate the shin inward. That inward rotation matters more than most people realize. It helps protect the knee from being forced into a knock-kneed (valgus) position, which is one of the most common mechanisms of ligament injury.

The Semimembranosus

Deeper and slightly behind the pes anserinus group sits the semimembranosus, the other medial hamstring. Unlike its neighbor the semitendinosus, which has a long thin tendon, the semimembranosus is more broadly anchored. Its distal tendon splits into five separate arms that weave into the back-inside corner of the knee, intertwining with the posterior oblique ligament to form a thick web of tissue that reinforces the posteromedial capsule.⁷ This branching design means the semimembranosus does not just bend the knee and rotate the shin. It also acts as a dynamic ligament, tightening the back-inside corner of the joint whenever it contracts.

The Adductor Magnus Tendon

One more structure worth mentioning sits at the very top of the inner knee. The adductor magnus, a large inner-thigh muscle, sends a tendon down to a bony bump called the adductor tubercle on the inner side of the femur, just above the joint line. In cadaver measurements, this tendon averages about 12.6 cm in length.⁸ The adductor magnus tendon matters clinically because it sits right next to the medial patellofemoral ligament (the ligament that prevents the kneecap from slipping outward), and surgeons sometimes use it as graft tissue when that ligament needs to be rebuilt. The VMO also takes some of its origin from the adductor magnus tendon, linking these two structures both anatomically and functionally.

How These Muscles Stabilize the Knee

The inner-knee muscles do far more than bend and straighten the leg. Collectively, they act as dynamic stabilizers, meaning they resist forces that would push the knee into abnormal positions during movement. The medial hamstrings (semitendinosus, gracilis, and semimembranosus) are especially important for preventing the knee from buckling inward. In biomechanical testing, simulating the pull of the semitendinosus alone significantly reduced valgus (knock-knee) angles at low flexion angles, bringing an injured knee back close to its intact values.⁹

This stabilizing effect is greatest between full extension and about 30 degrees of flexion, a range where the knee’s ligaments alone may not fully control side-to-side wobble, particularly if the posteromedial ligaments are damaged.¹⁰ The medial hamstrings also contribute to rotational and front-to-back control. Harvesting the gracilis and semitendinosus tendons for surgical grafts measurably alters the knee’s native movement patterns, increasing anterior translation, external rotation, and valgus motion under load.¹¹

Medial Knee Pain and What It Might Mean

If you feel pain on the inner side of your knee, the muscle and tendon anatomy described above helps narrow down the cause. One of the most common and commonly missed diagnoses is pes anserine bursitis, an inflammation of the fluid-filled sac that sits between the pes anserinus tendons and the shinbone. The pain typically shows up just below and slightly inside the joint line, and it often mimics a torn medial meniscus closely enough that unnecessary arthroscopic surgeries have been performed because the bursitis was missed on initial assessment.¹²

The key to distinguishing pes anserine bursitis from a meniscal tear or medial osteoarthritis is location. In meniscal tears and arthritis, tenderness sits right along the joint line itself. In pes anserine bursitis, the tenderness is below and slightly in front of that line.¹³ Runners, people with tight hamstrings, and those with obesity or diabetes are especially susceptible. Treatment usually involves rest, ice, and targeted stretching of the hamstrings and hip adductors.

Medial knee pain also correlates with broader joint health. In older adults, cartilage defects on the inner side of the tibial plateau carry significantly higher odds of knee pain, and lower quadriceps strength is an independent risk factor for both the pain and the underlying cartilage damage.¹⁴

The VMO, Kneecap Tracking, and a Persistent Myth

For decades, the rehabilitation world believed that a weak or shrunken VMO relative to the vastus lateralis was the primary culprit behind patellofemoral pain syndrome (PFPS), the diffuse aching around the front of the knee that accounts for a huge share of orthopedic visits. The logic was straightforward: if the VMO is too weak to counterbalance the outward pull of the vastus lateralis, the kneecap drifts laterally and grinds against the femur.

There is real biomechanical evidence behind this idea. In lab testing, progressively reducing VMO force caused significant increases in lateral patellar shift at low flexion angles, and complete VMO loss produced enough displacement to threaten subluxation.¹⁵ And in patients who show measurable patellar maltracking, delayed VMO activation correlates strongly with how far the kneecap drifts off course.¹⁶

But the story is messier than “weak VMO equals knee pain.” When researchers measured quadriceps muscle size in people with patellofemoral pain using MRI, one study found that VMO cross-sectional area was indeed smaller in the painful group compared to controls.¹⁷ However, a separate investigation found that the atrophy was not isolated to the VMO at all. All portions of the quadriceps were smaller on the affected side, and selective VMO wasting relative to the vastus lateralis could not be identified.¹⁸ A third study partially reconciled these findings by showing that VMO atrophy was more evident than vastus lateralis atrophy in the region immediately above the kneecap, while both muscles shrank further up the thigh. The researchers concluded that a combination of general quadriceps strengthening and targeted VMO work made sense for rehabilitation.¹⁹

The bottom line from the research is that the VMO clearly matters for kneecap tracking, but blaming patellofemoral pain on a weak VMO alone oversimplifies the problem. The whole quadriceps group tends to waste away when you have persistent knee pain, and simply hammering the VMO in isolation will not fix everything.

Can You Isolate the VMO During Exercise?

This is one of the most debated questions in knee rehab. Physical therapists have long prescribed exercises thought to preferentially target the VMO, such as short-arc terminal knee extensions, mini squats with the feet turned out, or straight leg raises with external rotation. The idea is to boost the VMO-to-vastus-lateralis activation ratio and restore balanced kneecap tracking.

An older electromyography study tested nine common exercise variations and found that none of them produced significantly greater VMO activity compared to the vastus lateralis, vastus intermedius, or vastus medialis longus. The conclusion was blunt: the VMO cannot be significantly isolated during these exercises.²⁰ For a long time, this was treated as the final word.

More recent work has pushed back, though. A 2024 study comparing a selective VMO strengthening program to a general quadriceps program found that the selective group achieved significant increases in VMO cross-sectional area and in the VMO-to-vastus-lateralis ratio for both muscle activation and size. The general group, by contrast, saw greater gains in the vastus lateralis.²¹ Proprioceptive neuromuscular facilitation (PNF) patterns have also shown promise: in people with patellofemoral pain, specific diagonal movement patterns produced higher VMO-to-vastus-lateralis ratios than simple straight leg raises.²²

The practical takeaway is that completely isolating the VMO the way you can isolate, say, a bicep is probably not realistic. But you can tip the balance in the VMO’s favor with the right exercise selection, and there is growing evidence that doing so improves the ratio in a clinically meaningful way. Most rehab programs now combine general quadriceps strengthening with exercises known to emphasize the VMO, rather than picking one approach over the other.

What Happens When Surgeons Harvest Medial Knee Tendons

The gracilis and semitendinosus tendons from the pes anserinus group are among the most commonly used grafts for anterior cruciate ligament (ACL) reconstruction. Because these tendons are accessible, appropriately sized, and expendable (other muscles can partially compensate for their loss), they have been a workhorse of knee surgery for decades. But removing them is not consequence-free.

At one year after surgery, both the semitendinosus and gracilis muscles show persistent atrophy on the operated side, and the semitendinosus muscle belly frequently retracts up the thigh. Hamstring strength deficits also persist at the one-year mark.²³ In systematic reviews comparing reconstruction using both tendons versus semitendinosus alone, harvesting the gracilis on top of the semitendinosus adds measurable weakness at deep knee flexion angles.²⁴ There is also evidence that losing the gracilis tendon specifically affects the shin’s ability to rotate inward at extreme angles.²⁵

These findings are why some surgeons now prefer using the semitendinosus tendon alone (quadrupled and folded) when possible, sparing the gracilis to preserve as much medial stability and rotational control as they can. The trade-off is a slightly thinner graft, so the decision depends on the patient’s body size, activity level, and the condition of their medial ligaments.

The Medial Gastrocnemius

One more muscle crosses the back-inside of the knee that often gets overlooked: the medial head of the gastrocnemius, the larger of the two calf muscle heads. It originates from the back of the femur just above the medial joint line and runs downward to merge into the Achilles tendon. Because it spans the knee joint, it is technically a knee flexor as well as a foot pusher. Its activation decreases when the knee is bent, which reduces the total force available for calf raises performed in a seated position compared to standing.²⁶ A strained medial gastrocnemius, sometimes called “tennis leg,” is another source of sharp medial-posterior knee pain that can be confused with a meniscal injury or a Baker’s cyst.

Exercise and Medial Knee Osteoarthritis

For people whose inner-knee pain stems from medial compartment osteoarthritis rather than a muscle or tendon issue, targeted exercise still helps, though the mechanism is different. A pilot study of patients with medial knee osteoarthritis found that a six-week exercise program significantly reduced pain and lowered the co-contraction of the vastus lateralis and biceps femoris muscles during walking. Co-contraction sounds like a good thing, but excessive simultaneous firing of opposing muscles compresses the joint surfaces together and increases load on already-damaged cartilage. Reducing that co-contraction appeared to be a key part of why the participants felt better.²⁷

Former athletes face a higher baseline risk: a meta-analysis pooling over 3,000 participants estimated that about 30 percent of former athletes develop knee osteoarthritis, a rate that likely exceeds the general population partly because of repetitive joint loading and prior injuries.²⁸ For this group, maintaining the strength of the medial muscles described throughout this article is not just about performance. It is about protecting the joint surface from further breakdown.