What Works
Tier 1 — Don't image it, and don't screen a tendon that doesn't hurt HIGH
This is the best-evidenced item on the page, and it is a decision not to do something. Nearly every pain-free knee looks abnormal on a scan, and when researchers scanned healthy athletes' tendons and followed them for years, only about one in four of those with an abnormal-looking tendon ever went on to develop pain.
The scan is very good at telling us a tendon is fine, and quite poor at telling us it's the cause.
Tier 2 — Progressive loading with real external weight MODERATE
The main treatment. Across 110 studies and nearly 4,000 people, two things stood out: adding external weight beat using bodyweight alone, and training less than daily beat training daily or twice daily. The number of reps mattered least of the three.
Slow heavy leg press or squat — 3 seconds down, 3 seconds up. 3-4 sets of 6-8. Three times per week, not daily.
Single-leg decline squat — slow descent on the painful leg, come up with both. 3 sets of 12-15, three times per week.
Split squat or step-down — controlled, adding weight as you progress. 3 sets of 8-10 each leg, two to three times per week.
Pain guide: up to 3-4 out of 10 during the exercise, settling within 24 hours, and no worse the next morning.
Tier 2 — Sort out kneecap tracking if the pain is above the kneecap MODERATE
People whose kneecap tracks slightly off to the outside had tendon changes in 57.8% of cases, compared with 27.3% of people whose kneecap tracked normally. If that is the driver, loading the tendon alone treats the symptom and leaves the cause running.
Tier 3 — optional and contested
Isometric holds for short-term pain relief LOW-MODERATE
Hard pushes against a fixed resistance, held still. Popular advice, and genuinely helpful for some people before training, but the claim that they beat normal lifting for pain relief came from studies of six and twenty people and did not survive when pooled with independent data.
Isometric knee extension hold — 5 holds of 30-45 seconds, hard effort, no movement. Before training or on painful days. Long holds and short holds work equally well when the total time adds up the same.
Decline squats, at any board angle MODERATE
The traditional 25 degree board was compared head-to-head against a 17 degree board and there was no difference on any measure. The angle was never the active ingredient, so pick whichever is comfortable.
Exercise Prescription
The full prescription sits inside the tiers above, next to the recommendation it belongs to. The short version: heavy, slow, three times a week, with real weight. Not bodyweight every day.
What Doesn't Work
- Shockwave therapy. When it was tested against a dummy version, with both groups also doing their exercises, the difference was 0.1 points on a pain scale. It survives on studies that never included a dummy comparison.
- Injections as standard care. No trial has compared them against a dummy injection, and the reported benefits are too large to be biologically believable for a pain treatment.
- Prevention programmes. Across 6,091 people they did not reduce the number of cases, and among military recruits the trend actually ran the wrong way.
- Surgery as a rescue. Compared against exercise at one year, it produced no meaningful advantage, and 85% returned to sport with exercise versus 86% with surgery.
- Prescribing reps precisely while leaving the weight vague. The classic "three sets of fifteen, twice a day, every day" pins down the least important variable and contradicts what the dosing evidence actually found about frequency.
- Treating a tendon because it looks thicker. In healthy young athletes with no pain at all, quadriceps tendons thickened by 7-11% over 1.7 years while their patellar tendons did not change. That is the tissue adapting, not breaking.
Red Flags — get seen urgently
- You felt or heard a pop, and now you cannot lift your straight leg off the bed or cannot straighten the knee properly. You may be able to feel a gap just above or just below the kneecap. That is a torn tendon, not tendon pain, and it needs same-day orthopaedic assessment. Do not load it.
- Both knees hurt, or the pain started without any increase in your activity.
- Morning stiffness lasting more than an hour, other joints involved, psoriasis, or long-standing back pain that is worse with rest and better with movement. That pattern points to an inflammatory condition rather than an overuse one, and it is treated completely differently.
- The knee is hot, very swollen, or the pain is out of proportion to what you actually did.
- You recently took a fluoroquinolone antibiotic (ciprofloxacin, levofloxacin) and the tendon pain is new. Speak to whoever prescribed it.
- You have known kidney disease and new tendon pain.
Refer to: Orthopaedics for a suspected torn tendon (urgent). Your doctor or a rheumatologist for inflammatory or systemic features. Emergency care if a joint infection is possible.
Point at your pain with one finger. If it lands on a sharp, specific spot right at the top or the bottom edge of your kneecap, and it is worst when you jump, land or squat deep, you are most likely dealing with the tendon. If instead you find yourself covering the whole kneecap with your palm, that is a different problem, and this is not the protocol you need.
Load it heavier and less often. Don't rest it, and don't get it scanned.
Your kneecap works like a pulley, with a cable running in above it and out below it. Tendon pain is not the cable snapping. It is the cable being asked to carry more each week than it can rebuild between sessions. That is why total rest does not fix it and hammering it every day does not either: the rebuilding only speeds up when you load it hard enough to trigger repair, but leave it long enough to actually finish.
- What this actually is: not inflammation, and not damage that needs protecting, but a tendon whose capacity has quietly fallen behind what you are asking of it.
- What most people get wrong: resting it and scanning it. In one study of 230 knees belonging to people with no pain whatsoever, 97% showed something abnormal on the scan.
- The one change that matters: load it with real weight, slowly, about three times a week, instead of a little bit every single day.
Best for
Front-of-knee pain that is focal, sits at one edge of the kneecap, and clearly tracks with load. Whether you are a jumper, a lifter, or someone who just started running again.
Skip if
You cannot lift your straight leg, both knees are affected, or you have inflammatory or systemic symptoms. Those need proper assessment first, and a physical therapist is the person who sorts that out quickly.
Want the full evidence? Keep scrolling
When you're actually ready
- Full knee movement, with no pain
- Single-leg decline squat pain of 2/10 or less, at the depth your sport or lift actually needs
- Knee extension strength within 10% of the other side
- Three sets of eight at your previous working weight, controlled, with no flare the next day
- Symptom-free during and 24 hours after two consecutive weeks of full training loads
- Jumping and landing rebuilt gradually, with no pain on landing (jumping sports only)
- No pain on stairs or deep squatting in ordinary daily life
How confident is this? Moderate
The unusual thing about this topic is that the strongest claims are claims about what is missing.
It is close to certain that quadriceps tendinopathy has no treatment trials of its own, and close to certain that scans cannot tell you where your pain is coming from. It is only moderately certain that loading is the right first treatment, because the trials behind it are small, unblinded, and have never once been compared against a placebo.
What would change the loading recommendation?
A properly sized trial (200 people or more) comparing real loading against a convincing dummy exercise programme, matched for time, supervision and attention, measuring outcomes at a year. If loading won clearly, this moves to high confidence. If it did not, then loading stays the sensible first choice on grounds of safety and cost rather than proven effect. That trial has never been run.
What would change the quadriceps tendon recommendation?
A single decent trial of 80 or more people with quadriceps tendon pain specifically, split by whether their kneecap tracks normally, comparing loading against usual care over six months. One trial would convert this whole half of the page from reasoned extrapolation into actual evidence. It would also be the first of its kind.
Don't want to guess which knee pain needs loading and which needs a referral?
The Full Picture — Anatomy, Diagnosis & Evidence
What's Actually Going On
Your quadriceps muscle runs into a tendon that attaches to the top of your kneecap. The kneecap acts as a pulley, and a second tendon carries the force from the bottom of the kneecap down to your shin bone. Together they are called the extensor mechanism, and it is what straightens your knee and controls you on the way down.
Force concentrates where tendon meets bone, which is why pain shows up at the top or bottom edge of the kneecap rather than in the middle of the tendon.
This is a wearing-down problem, not an inflammation problem. The clearest evidence comes from a study of 891 tendons that had ruptured on their own: not a single one had healthy tissue underneath, while two thirds of comparison tendons did. The wearing-down comes first, quietly, and it is common with age.
An honest caveat about the mechanism
Clinicians usually explain loading by saying it remodels the tendon. When researchers actually tested that on 76 athletes, the benefit of a loading programme was not explained by changes in strength, ankle movement, jump performance, tendon thickness, or blood vessel growth. Every one of those came back with no measurable link. Loading appears to help. Nobody has yet demonstrated how.
How to Identify It
The most useful thing you can do costs nothing: point at it with one finger.
- Top edge of the kneecap — the quadriceps tendon. More common in older people, lifters, and those whose kneecap tracks slightly outward.
- Bottom edge of the kneecap — the patellar tendon. The classic jumper's problem, and the one nearly all the research is about.
- Vague ache behind or all around the kneecap — probably not a tendon at all. That pattern usually points to patellofemoral pain, which is provoked by long periods sitting rather than by jumping and landing.
- Warms up, then hurts more afterwards — very typical of tendon problems.
- Swelling of the whole joint, night pain, or rest pain — not typical. Look elsewhere.
Why there are no accuracy numbers on this page
You will often see confident percentages quoted for how well the hands-on knee tests perform. We searched an evidence base of 83 papers and found no study establishing how often any of these tests correctly catches or correctly clears either tendon problem. Not for the decline squat, not for pressing on the tendon, not for resisted straightening.
That does not mean the tests are useless. They are how a clinician locates the problem and tracks whether it is improving, and the decline squat is used as the standard pain measure throughout the research. It means something narrower: nobody has measured how reliably they distinguish this from everything else. Anyone quoting you a percentage should be able to name the study it came from.
The Debate
There is no clinical guideline for either of these conditions. Not an outdated one, not a disputed one. None. One of the reviews we retrieved says so outright. So the disagreements here are between what is widely taught and what the trials actually show.
Is loading a well-established treatment?
Does shockwave therapy help?
How should the exercises actually be dosed?
Honest Limitations
The people in the studies are probably not you
The finding: loading improves patellar tendon pain.
The gap: every trial in the 2025 Cochrane review enrolled athletes, 88% of them male, average age 26, who had already been in pain for over three years. The review itself says this limits the findings to athletes rather than the general public. The person most likely to get quadriceps tendon pain is older, often has a kneecap that tracks off-line, and appears in none of these trials.
What to do with that: treat on principles, and do not expect an athlete's timeline if you are not one.
Half of this page is an extrapolation, and you should know which half
The finding: there is a body of evidence for treating the tendon below the kneecap.
The gap: for the tendon above the kneecap, there is no trial evidence at all. Across 83 papers, the only treatment report specific to it is a single case report about one person, whose authors noted it was the first ever published.
What to do with that: the plan is the same, because it is the best available reasoning. But it is reasoning, not proof, and you deserve to be told which one you are getting.
Time is doing some of the work, and nobody knows how much
The finding: people improve on loading programmes, and most report being satisfied.
The gap: untreated tendon pain also improves somewhat on its own, satisfaction rises the longer you follow people up, and no trial has ever included a convincing dummy exercise group. So the share of the benefit that comes from the loading itself, rather than from time, attention and expectation, has never been measured.
What to do with that: this is a reason to be humble about the explanation, not a reason to skip the treatment. Waiting is not the better option either, since untreated tendon pain did not fully resolve within three to four months.
The Nuance
On surgery. In the one trial that compared them directly at twelve months, 85% returned to sport after exercise and 86% after surgery. A Cochrane review of surgery found no clinically important benefit over exercise on pain, function, or overall improvement, and no trial has ever compared surgery against a placebo operation. Given that surgery carries risk and exercise does not, the burden of proof sits with the operation. A genuinely ruptured tendon is a completely different situation and does need surgical assessment.
On the outcome scores. The questionnaire used throughout this research is reliable in the sense that it gives consistent answers, but a formal review rated the evidence for whether it measures the right things as very low quality. There is also no version of it validated for the quadriceps tendon at all. And no study has established how big a change on it has to be before it actually means something to a patient, which is why you will not see a threshold quoted here.
On strength. People with patellar tendon pain do measurably lack knee extension strength compared with people who do not have it. But because those studies compared groups at a single point in time, nobody can say whether the weakness caused the pain or the pain caused the weakness. Worth training either way. Not worth treating as proof of the mechanism.
Key References
- Lopes AD, Rizzo RR, Hespanhol L, Costa LO, Kamper SJ (2025), Cochrane Database of Systematic Reviews. Exercise for patellar tendinopathy. Seven trials, 211 participants; low to very low certainty; no placebo comparison exists. PMID 40421598
- Dan M, Phillips A, Johnston RV, et al. (2019), Cochrane Database of Systematic Reviews. Surgery gave no clinically important benefit over eccentric exercise. Two trials, 92 participants. PMID 31546279
- Pavlova AV, Shim JSC, Moss R, et al. (2023), British Journal of Sports Medicine. Resistance exercise dosing across 110 studies and 3,953 participants. PMID 37169370
- Challoumas D, Pedret C, Millar NL, et al. (2021), BMJ Open Sport and Exercise Medicine. Network analysis of 37 randomised trials; shockwave no better than a dummy version. PMID 34900334
- Horga LM, Hirschmann AC, Henckel J, et al. (2020), Skeletal Radiology. 230 knees of asymptomatic adults; 97% showed an abnormality. PMID 32060622
- Cushman DM, Vomer R, Teramoto M, et al. (2025), Clinical Journal of Sport Medicine. 1,156 pain-free patellar tendons followed over time. PMID 38864880
- Deng J, Runhaar J, Breda SJ, et al. (2025), Journal of Science and Medicine in Sport. Loading benefit was not explained by any physical or imaging change. PMID 39718487
- Ghany JF, Kamel S, Zoga A, et al. (2021), Skeletal Radiology. Extensor tendon change in 57.8% of kneecap maltrackers versus 27.3% of controls. PMID 33876276
- Visnes H, Tegnander A, Bahr R (2015), Scandinavian Journal of Medicine and Science in Sports. Quadriceps tendon thickened 7-11% in healthy athletes. PMID 24612006
- Kannus P, Jozsa L (1991), Journal of Bone and Joint Surgery. 891 spontaneously ruptured tendons; none had healthy tissue. PMID 1748700
- Obst SJ, Peterson B, Heales LJ (2024), Journal of Athletic Training. Knee extension strength deficits across 23 studies. PMID 37071550
- Wang S, Lyu B (2023), Sports Health. Prevention programmes across 6,091 participants; no significant reduction. PMID 36146934
- Garzon M, Balasch-Bernat M, Cook C, et al. (2024), Musculoskeletal Science and Practice. Natural history of untreated tendinopathy. PMID 38879981
- Korakakis V, Kotsifaki A, Stefanakis M, et al. (2021), Knee Surgery Sports Traumatology Arthroscopy. Outcome questionnaires rated very low quality for content validity. PMID 34019117