Yoga for knee pain: what the trials show, and how to protect your knees

Knee arthritis: a modest benefit

Across 8 trials (756 people), yoga eased knee arthritis pain and stiffness and helped people get about, compared with control groups, and a review of 14 studies agreed, on low-certainty evidence.

Sources 1, 2

As good as strengthening, in the best trial

In 117 adults in Australia, 12 weeks of yoga eased knee pain as much as a strengthening programme, and it was likely to be cost-effective. It didn't do better than strengthening.

Sources 3, 4

Classes did better than videos

Unsupervised online yoga videos gave a small gain in function that had gone by 24 weeks, while the trials with teachers did better.

Sources 5, 3

Guidelines count it as exercise

The American College of Rheumatology conditionally recommends yoga for knee arthritis. NICE (the National Institute for Health and Care Excellence) recommends exercise tailored to you without naming a type, and its evidence review found no difference between types.

Sources 6, 7, 8

Bend only a little, and pad your knees

NCCIH (the US National Center for Complementary and Integrative Health) names the knee or lower leg as the part most often injured in yoga. In a lab study, a wide squat and an upright lunge put the least load on the inside of the knee, and the Arthritis Foundation advises people with arthritis not to bend their knees past a right angle.

Sources 9, 10, 11

Some knee pain needs checking first

Get advice from NHS 111 for a knee that is hot, red or swollen with a high temperature, or one that locks, gives way or clicks painfully. The full list is in the warning signs below.

Source 12

Which knee pain the trials studied

Every randomised yoga trial we found for knee pain was in knee osteoarthritis, the most common kind of knee arthritis, mostly in adults aged 50 to 80 and mostly in women. We found no randomised yoga trial in kneecap pain or a torn meniscus (a PubMed search, October 2026).

Sources 13, 1, 2, 14

Why it might help

Yoga combines poses that strengthen and stretch the legs with breathing and relaxation. The trial authors suggest it may ease pain through strength, flexibility, less stress and better body awareness, but how it works hasn't been tested.

Sources 3, 15

Kneecap pain is a different problem

Kneecap (patellofemoral) pain is pain beneath or around the kneecap, "especially on climbing up and down stairs, kneeling and squatting". NHS inform says the strongest evidence is for strengthening weak or underactive muscles and stretching short or tight ones.

Source 16

Knee arthritis: modest benefits, low certaintyEmerging evidence

Across 8 trials (756 people), yoga eased pain and stiffness and helped function compared with control groups, though results varied widely between trials, and it didn't clearly improve daily activities or quality of life. A 2026 review of 14 studies (1,183 people) found the same on low-certainty evidence, with the most pain relief from programmes of 4 or more sessions a week for at least 8 weeks. An earlier review of 9 trials rated all its evidence very low quality.

Sources 1, 2, 13

As good as strengthening exerciseEmerging evidence

In 117 adults with knee arthritis in Tasmania (average age 62), two 1-hour supervised classes and one home session a week for 12 weeks eased pain by about 18 points out of 100, against about 17 with a strengthening programme. That isn't better, but it is within the margin the trial set for "as good as". By 24 weeks, with home practice only, knee pain, function and stiffness scores favoured yoga. Yoga was likely to be cost-effective.

Sources 3, 4

Side effects in that trialEmerging evidence

Side effects possibly or probably caused by the exercise were more common with yoga (15.5% of people) than with strengthening (6.8%), including more knee pain in 6 people against 5. None were serious. The authors suggest unfamiliar poses may explain the difference.

Source 3

Against aerobic and strength exercise in older adultsEmerging evidence

In 83 older adults, 8 weekly 45-minute hatha classes with home practice improved knee symptoms, anxiety and fear of falling more than aerobic and strengthening classes, and the 3 minor side effects were all in the exercise group. In 36 older women, 8 weekly classes eased pain and stiffness against a waiting list. Both were small pilot trials.

Sources 17, 15

Online videos on their own did lessEmerging evidence

In 212 adults, 12 weeks of unsupervised online yoga videos improved function by 4 points on a 68-point scale, less than the change people notice, and didn't clearly ease walking pain. Both had faded by 24 weeks.

Source 5

Other trialsMixed evidence

In India, 250 adults who had electrotherapy and then 40 minutes a day of supervised yoga or exercise for 3 months improved more with yoga. In 75 inactive adults with arthritis, 8 weeks of yoga improved physical health, walking and mood against a waiting list. In 131 older adults who couldn't do standing exercise, chair yoga twice a week for 8 weeks eased how much pain got in the way of life, and that lasted 3 months.

Sources 18, 19, 20

Rankings put other mind-body exercise higherMixed evidence

Network analyses rank exercise types partly through indirect comparisons. One of 39 studies ranked yoga first for stiffness and function; one of mind-body exercise ranked Pilates and tai chi first, with yoga's effects on low-certainty evidence; and a third ranked tai chi and Wuqinxi (a Chinese exercise based on animal movements) first for pain, with yoga helping mental health. A ranking is not proof that one exercise beats another.

Sources 21, 22, 23

American College of RheumatologyStrong evidence

Its 2019 guideline with the Arthritis Foundation strongly recommends exercise and tai chi for knee arthritis, and conditionally recommends yoga, which it calls "far less well studied than tai chi". It made no recommendation on yoga for hip arthritis, "due to lack of data".

Source 6

NICEStrong evidence

NICE says to offer everyone with osteoarthritis "therapeutic exercise tailored to their needs (for example, local muscle strengthening, general aerobic fitness)", and to consider supervised sessions. It names no type of exercise. Its evidence review grouped the yoga trials with mind-body exercise, and the committee "noted that there did not appear to be a difference between different types of exercise".

NICE also says joint pain may increase when you start exercising, and that regular, consistent exercise still helps your joints.

Sources 7, 24, 8

OARSIStrong evidence

The Osteoarthritis Research Society International (OARSI) makes education and structured land-based exercise the core treatments for knee osteoarthritis.

Source 25

The NHSStrong evidence

The NHS calls exercise "one of the most important treatments for people with osteoarthritis", combining strength and general fitness. It adds that doing too much too quickly, or the wrong sort of exercise, "may damage your joints", so follow a plan from your GP or physiotherapist. Our knee arthritis guide notes that no exercise has been shown to damage the joint. The practical advice is the same either way: build up gradually.

Source 26

A Canadian panelEmerging evidence

The Ottawa panel graded hatha yoga B for knee arthritis pain and C+ for function, from a small number of trials.

Source 27

What this means for you

Yoga counts as exercise for knee arthritis. It isn't better than strengthening exercise, and strength work for your thighs and hips is the backbone, so pair yoga with it.

Sources 6, 7, 3

Hatha classes for older adults

Cheung's trials ran one 45 to 60-minute class a week for 8 weeks, plus 30 minutes at home 4 times a week, in small classes (9 people in the first trial) with a mat, blocks, a strap, blankets and a chair. Poses were changed to suit each person. The key poses were:

• mountain pose, warrior I and II, and a tree pose variation
• chair pose
• easy seated pose, bound angle and open angle
• a half locust variation and bridge
• a standing forward fold and head-to-knee pose
• a lying hamstring stretch with a strap, a lying twist and lying bound angle
• relaxation pose

Sources 15, 17

Home practice is hard to keep up

In the first trial, only a third of people did the home practice as prescribed, though most kept practising at home 20 weeks on.

Source 15

The Tasmanian trial's classes

Two 1-hour classes a week, led by yoga teachers or physiotherapists in groups of 10, plus one session at home, for 12 weeks, then 3 home sessions a week. Each class opened with 10 minutes of breathing and chanting, then 40 minutes of warm-up and poses, and ended with 10 minutes of deep relaxation and meditation, with the practices "modified for the needs and limitations of each participant".

Sources 3, 28

Knee-friendly squats and lungesEmerging evidence

In a lab study of healthy young women, a wide squat (goddess) and a lunge with the body upright (warrior) put the least load on the inside of the knee, and a one-leg balance the most. A Canadian team built a programme from squats and lunges, bending the knees to 30 degrees at first, then 60, then 80, with the arms forward and then overhead, 3 times a week for 12 weeks.

In 45 women with knee arthritis it eased pain and built strength, with knee loads in the poses below those of walking, but 2 of the 45 developed kneecap pain. In a small trial of 31 women it did better than no exercise and as well as machine strengthening.

Sources 10, 29, 30

Classes rather than videos

Unsupervised videos gave small, short-lived gains, and NICE's committee judged supervised exercise likely to help more than unsupervised exercise. The Arthritis Foundation suggests starting with a class or a private teacher who can help you change poses, and telling them about your arthritis.

Sources 5, 24, 31

What the trials usedEmerging evidence

Two supervised 1-hour sessions plus 1 home session a week for 12 weeks in the best trial; one class a week plus 30 minutes at home 4 times a week for 8 weeks in older adults; and three 1-hour classes a week for 12 weeks in the knee-friendly programme. Across 14 studies, 4 or more sessions a week for at least 8 weeks gave the most pain relief.

Sources 3, 15, 30, 2

The pain rule

Our knee guides use this rule: rate your knee pain from 0 to 10 during and after exercise; 0 to 2 counts as safe, up to 5 as acceptable and above 5 as too much, and your knee should be back to its usual level by the next morning. In the knee-friendly yoga trial, pain was allowed to rise by no more than 2 points in a session. NICE says joint pain may rise when you start exercising with arthritis.

Sources 30, 7

Our editorial starting point

Two or three sessions a week of 25 to 45 minutes, for at least 8 to 12 weeks, ideally starting with a supervised class, plus strength work for your thighs and hips on 2 or 3 days. Bend a little deeper over the weeks, and with arthritis, never past a right angle.

Sources 3, 2, 7

Kneecap pain and runner's kneeMixed evidence

We found no randomised yoga trial for kneecap pain, only one uncontrolled study: 24 teenagers with front-of-knee pain followed yoga videos 3 to 4 times a week for 8 weeks, flexibility first and then strength, and their knee scores improved. Strengthening and stretching have the strongest evidence, and NHS inform suggests regular breaks from the activities that bring it on. Yoga can provide stretching and some strength if you skip kneeling and deep squats while it's sore.

Sources 32, 14, 16

A torn meniscusMixed evidence

Yoga hasn't been tested for a torn meniscus, the cartilage cushion inside the knee. People who do yoga have more meniscus injuries than people who don't, in a review of observational studies, and meniscus tears show up among yoga injuries on scans. Avoid lotus, hero pose, deep squats and twisting on a bent knee.

Sources 33, 34, 35

After a knee replacement

Follow your physiotherapy exercises first. The NHS says "do not kneel on your new knee until your doctor says you can" and "do not sit with your legs crossed for the first 6 weeks". In a survey, 120 European knee surgeons agreed that yoga could be allowed from 6 to 12 weeks after surgery; ask your own team. One small study added yoga poses to physiotherapy, but its design was weak.

Sources 36, 37, 38

Older adults and chair yoga

Chair yoga eased how much pain got in the way of daily life in older adults with leg arthritis who couldn't do standing exercise. In the US, people aged 65 and over had the highest rate of emergency visits for yoga injuries, so keep a chair or wall close in standing poses. Our yoga for seniors and chair yoga guides have more.

Sources 20, 39

Where yoga injuries happenEmerging evidence

NCCIH says "the parts of the body most commonly injured are the knee or lower leg". US emergency department records of 29,590 yoga injuries put the trunk first (about 47%), so the knee is a common site rather than always the main one.

Sources 9, 39

Kneeling

Pad your knees with a folded mat or blanket, keep kneeling short, and swap the pose if it hurts. Kneeling often aggravates kneecap pain. Come out at once if your foot or lower leg goes numb or tingles: a nerve injury causing foot drop (weakness lifting the foot) has been reported after a kneeling yoga pose. Our knee pain when kneeling guide explains why everyday kneeling hasn't been shown to damage knees.

Sources 40, 16, 36

Deep bendsEmerging evidence

Rising from a full squat or from kneeling put about 4,500 newtons through healthy knees in a lab study, several times body weight. The Arthritis Foundation advises people with arthritis to avoid poses that bend the knees more than 90 degrees, and adds: "Don't do anything that hurts."

Sources 41, 11

One-leg balancesEmerging evidence

A one-leg balance put the most load on the inside of the knee in the lab study, and the Arthritis Foundation advises people with arthritis to avoid poses that balance on one foot. In tree pose, keep the toes of your lifted foot on the floor and a hand on a wall, or skip it.

Sources 10, 11

Lotus and hero pose

NCCIH advises beginners to avoid the lotus position, and lotus and half lotus caused 5.3% of injuries in an Australian survey. In hero pose, sit high on a block or leave it out, and skip it after meniscus or ligament surgery or if your kneecap hurts (the move page's caution).

Sources 9, 42

Props

The trial classes used blocks, straps, blankets and a chair. A cushion under your hips when sitting, padding under kneeling knees, a strap for hamstring stretches and a wall or chair for balance all take load off the knee (our suggestion).

Sources 15, 31

Get urgent medical help if

These are the same warning signs as our knee pain guides. In the US, call 911 for 999, and use your doctor's urgent line or urgent care for NHS 111.

• Leg pain or swelling with breathlessness or chest pain: call 999 or go to A&E, and do not drive yourself. It can be a blood clot that has reached the lungs.
• Back or leg symptoms with numbness around the genitals or bottom, new trouble controlling your bladder or bowels, or severe or worsening weakness or numbness in both legs.
• Get advice from NHS 111 now: a hot, red or swollen knee with sudden severe pain, a very high temperature, or feeling hot, cold or shivery. This can be a joint infection.
• You cannot move the knee or put any weight on it, or it is very painful, badly swollen or has changed shape.
• You hurt your knee and cannot take 4 steps, or you are 55 or over and hurt it in a fall: get it checked the same day for a broken bone.
• Throbbing pain and swelling in one calf or thigh with red or darkened skin or swollen veins, or sudden worse pain and swelling around a lump at the back of the knee (a possible clot or a burst Baker's cyst).
• The knee locks, gives way or clicks painfully. Painless clicking is normal.

Source 12

See a doctor or physiotherapist if

• Knee pain that has not improved within a few weeks of looking after it at home. Many areas offer NHS physiotherapy without a GP referral.
• Numbness, tingling or weakness that travels down the leg from the back. That points to a nerve (sciatica) rather than the knee.

Source 12

During yoga

Sharp, catching or giving-way pain, or new swelling, is a reason to stop and get it checked. Come out of a pose at once if your foot or lower leg goes numb or tingles in a kneeling or cross-legged pose, or if you feel a sharp pain inside the knee when twisting or rising from a deep bend (our suggestion, from case reports and injury studies).

Sources 40, 33

After a knee replacement

Follow your surgical team's advice and the NHS: "do not kneel on your new knee until your doctor says you can". Our knee replacement guide lists the signs of infection or a clot that need urgent help.

Source 36

Tell the teacher

The Arthritis Foundation suggests arriving early to tell the teacher about your arthritis and ask which poses to change. Do the same for a knee replacement or a meniscus problem.

Source 31

"Yoga is bad for your knees"Mixed evidence

NCCIH names the knee or lower leg as the part most often injured in yoga, and meniscus injuries are more common in people who do yoga. But in knee arthritis trials, side effects were mild and none were serious. The risk sits with deep bends, lotus and kneeling, which are easy to change.

Sources 9, 33, 3, 17, 30

"With bad knees you shouldn't bend them"Mixed evidence

Knee-friendly yoga programmes bent the knees to 30, then 60, then 80 degrees, and eased pain and built strength. The Arthritis Foundation's limit for people with arthritis is a right angle, not no bending at all.

Sources 29, 30, 11

"Yoga is better than strengthening for knee arthritis"Mixed evidence

In the best trial yoga matched strengthening; it didn't beat it. NICE found no clear difference between types of exercise, and one network analysis ranked Pilates and tai chi above yoga.

Sources 3, 8, 22

"Yoga videos are as good as a class"Mixed evidence

Unsupervised online yoga gave a small gain in function that had faded by 24 weeks. Supervised classes did better, and NICE's committee judged supervised exercise likely to help more.

Sources 5, 3, 24

"Yoga can reverse knee arthritis"Mixed evidence

The trials measured pain, stiffness and function, not joint repair. Exercise is still "one of the most important treatments" for osteoarthritis, and the NHS advises building up gradually with a plan.

Sources 15, 3, 26

"Never kneel with knee pain"Mixed evidence

Kneeling often aggravates kneecap pain, and you shouldn't kneel on a new knee until your doctor says you can. Our kneeling guide explains why everyday kneeling hasn't been shown to damage knees, so pad it, shorten it or swap the pose.

Sources 16, 36