"Go for a swim, it is the only sport that does not wreck your joints." It is perhaps the most repeated advice in physiotherapy clinics, and it is not wrong. It is, however, more fragile than it sounds.
The research of the last twenty years says two things at once: water really does help people with knee pain, and it does not help more than a good programme on land. It is worth knowing this before buying a pool membership, because it changes your expectations and changes what to do when the results do not come.
In this article you will find what is documented about the relationship between swimming and the knee, with the real numbers and the stated certainty of the evidence: how much load water really removes (far less than the 90% you read everywhere), for whom the pool is the right choice and for whom it is a fallback, which stroke suits best and how to combine water with work on land.
Important before continuing. The exercises and advice that follow are general indications and may not suit your situation. Everyone has a unique clinical history: what works for others could be counterproductive for you.
If you want a professional's opinion, you can start with the questionnaire with our physiotherapists and assess your case together.
First of all: when the knee should not go in the pool
Water is gentle, but it is not a diagnosis. There are conditions in which the problem needs to be looked at before it is trained, and in which postponing a visit in order to go swimming wastes precious time.
- A knee that is hot, swollen and red, perhaps with fever or malaise: it may be acute inflammation or a joint infection, and it is a situation to assess promptly. We discuss it in detail in the article on a hot, swollen knee.
- Swelling that appears within a few hours of a trauma, with the sensation that the knee does not hold: it needs assessing before any activity.
- Joint locking: the knee stays bent and does not fully straighten, or it catches in jerks during movement.
- Sudden giving way while walking, with no pain that warns of it beforehand.
- Pain that wakes you at night or is present at rest, that never changes in any position, regardless of what you do.
- A calf that is swollen, warm and painful on one side only, especially after immobilisation or a long journey: it is a suspected vascular problem, and it is not trained, it is seen by a doctor.
None of these situations improves because you get in the pool, and some get worse if you wait. Swimming for the knee makes sense when you know what you are training.
What the studies really say about swimming and the knee
Here comes the uncomfortable part, and we write it anyway, with the numbers and with the certainty that the authors themselves declare.
The best evidence concerns exercise in water (hydrokinesitherapy, therapeutic water aerobics, walking in the pool) in people with knee and hip osteoarthritis.
The Cochrane review that pooled 13 studies and 1,190 participants concludes that, compared with doing nothing, exercise in water produces small and short-term effects: about five points less pain on a 0 to 100 scale (confidence interval from 3 to 8 points) and a similar reduction in disability.
Two details that almost nobody reports. The first: the certainty of the evidence is moderate, not high, and the authors describe the effects as small but clinically relevant. The second: none of the included studies repeated X-rays, so there is not even data to say that water changes anything in the joint. It reduces symptoms, it does not regenerate cartilage.
A more recent meta-analysis of 22 studies then asked the decisive question: is water better than land? In the 13 studies that directly compared exercise in water with exercise on land, about 650 participants, the differences in pain, stiffness and function were not significant.
Water does not beat the gym, the mat or the stairs: it ties. And of the benefits, three months after the end of the programme, only the one on pain remains measurable.
And how much is land-based exercise worth, in numbers
For fairness this needs saying too, so that the comparison is honest. The 2024 Cochrane update on land-based exercise pooled 139 studies and 12,468 participants. Compared with no treatment or usual care, exercise improves pain by about 13 points out of 100 (low certainty) and physical function by about 12.5 points (moderate certainty).
And here comes the part that makes the rest credible: the authors compared those numbers with the minimal important difference, that is, the change a person actually perceives (12 points for pain, 13 for function), and concluded that the clinical importance of these benefits remains uncertain.
This is not a rejection of exercise, which remains the first choice: it is the real measure of what to expect.
Be careful not to do the wrong sum, a mistake that is often seen: you cannot subtract the 5 points from water from the 13 points from land. These are different reviews, with different studies, populations and comparisons. The only legitimate comparison is head to head, and that is the one that says a tie.
What the guidelines say
International recommendations are consistent with one another. The OARSI 2019 guidelines identify education and structured land-based exercise programmes as the fundamental treatment for knee osteoarthritis, with or without weight management through diet.
Exercise in water, in the same guidelines, appears at a lower level of recommendation for the knee, depending on the person's other health conditions, and it is not recommended for hip or polyarticular osteoarthritis. In short: water is a good supporting option, not the first choice.
The US ACR 2019 guidelines are in line, giving a strong recommendation to exercise in general and, for people with knee or hip osteoarthritis who are overweight, to weight loss.
And real swimming, lengths of the pool
On actual swimming, with strokes rather than exercises in the pool, there is only one reference randomised trial: 48 sedentary middle-aged and older adults with osteoarthritis.
Three months of supervised swimming, 45 minutes three times a week at 60-70% of heart rate reserve, reduced pain and stiffness, increased strength (with gains of 15-30% in knee extension and flexion power) and lengthened the distance covered in six minutes.
An excellent result, with a detail that changes the reading: the comparison group cycled, and achieved exactly the same improvements. It is a small study, on people who started out sedentary, so it should be taken for what it is: a good clue, not a definitive proof.
Swimming or doing exercises in water: not the same thing
This distinction is missing from almost every article on the subject, and it is the most useful of all. When reviews talk about exercise in water, they almost never mean front crawl lengths: they mean walking in the pool, knee flexion and extension against the resistance of the water, structured programmes in the pool.
Swimming is a cyclical, aerobic action: it trains the breath, requires technique, almost entirely unloads the knee and shifts the work onto the shoulders. If you do not have a decent stroke, you tire without training your legs.
Exercising in water is something else: you dose the difficulty by choosing the depth and the speed of movement, you work on the strength and control of a specific joint, and you do not need to know how to swim.
If the problem is the knee, this route is more targeted, and it is also the one the studies have tested most. We discuss it in physiotherapy in water.
In practice: swimming is excellent for moving, for the breath and for the mind; exercises in water are the tool when the target is a specific knee.
The myth of 90% less load
You find it written everywhere, and it has a distorted grain of truth. A study of 193 healthy young adults measured the weight actually taken off the legs in the pool, with immersion up to the anterior superior iliac spines, the two bony prominences at the front of the pelvis.
Standing upright, the load was still about 52% of body weight, and dropped to around 43% when leaning forward with support. So at that depth you are still bearing about half of yourself, not 10%.
That study measured only one immersion height, so it does not tell you how much is unloaded at greater depth: by the principle of buoyancy, the more water you displace the less you weigh, and the load really drops only well above the pelvis. It should also be added that these were healthy young people, not patients, but the order of magnitude remains indicative.
The practical point is simple: when talking about swimming and the knee, the depth of the water is the dose. If you have been told to start by unloading the knee and you walk in shallow water, you are doing an exercise far more demanding than you think. And to make an exercise easier, it is often not necessary to change the exercise: just move to where the water is deeper.
Why always unloading the knee is not a goal
This is the point where the common belief does the most damage. "Zero impact" sounds like an absolute quality, and it is not: load is not the enemy of the joint, it is the signal that keeps it healthy.
A systematic review gathered the randomised trials that looked at cartilage with magnetic resonance imaging in people with knee osteoarthritis or at risk of developing it. Conclusion: exercise with load on the joint does not seem to be harmful to cartilage.
The certainty of the evidence is low and the studies are few, but the direction is clear and runs opposite to the most widespread fear.
From here everything else follows. The useful question is not "how do I avoid load", but "how much load do I tolerate today, and how do I increase it without triggering pain". Water serves to keep you moving while that tolerance is rebuilt, not to replace it.
In life the knee has to bear the weight of the body: stairs, pavements, slopes, getting up from a chair. In water that weight is absent, and that is exactly why it does not hurt, but it is also why the pool alone does not train the capacity to bear load. Going down stairs is only trained by going down stairs, in tolerable doses.
That is why the guidelines put land-based exercise first, and it is why the most useful sentence is also the simplest one: water is a phase, not a destination. Swimming is an excellent bridge, and a bridge has to lead somewhere.
For whom swimming is the right choice, and for whom it is a fallback
Here is the practical part, which is worth more than any ranking of sports.
Swimming is a good choice when:
- pain prevents you from bearing load: walking for ten minutes hurts, but in the pool you can move for forty without paying for it the next day;
- you are in an acute phase or a flare-up and the realistic goal is not to improve but not to stop completely;
- you have significant excess weight and on land you struggle to build up activity volume, while in water you manage it from the start;
- you are in the post-operative phase and the surgeon has given the go-ahead, with the wound completely closed;
- you are afraid of movement after a painful episode, and in water you feel secure enough to really move;
- you enjoy it: adherence over time beats any theoretical optimisation, and the best programme is the one you keep doing for months.
Swimming is, by contrast, a fallback when:
- on land you could train with acceptable discomfort, and you go to the pool only out of caution: in that case you are choosing the longer route;
- your goal is to get back to walking in the mountains, running or simply taking the stairs without thinking, because those abilities are trained under load;
- the main problem is the strength of the quadriceps and the glutes, which the pool trains little and poorly compared with targeted work;
- the stated aim is weight loss but you swim for an hour or so a week at an easy pace;
- you are using it to postpone an assessment, while the knee is giving signals that deserve a diagnosis.
Swimming and the knee: which stroke to choose between crawl, backstroke and breaststroke
Freestyle and backstroke are the gentlest on the knee: the kick is a flutter with a slightly bent knee, without rotation and without load. Those with knee osteoarthritis or recovering from a joint problem generally tolerate them well from the start.
Breaststroke is the one to be aware of. The whip kick closes and opens the knee with the shin rotated outwards, at high speed, many times per length. Among competitive swimmers there is a picture described for decades, "breaststroker's knee": pain on the inner side of the knee.
One historical study examined nine competitive swimmers with this pain arthroscopically and found, in seven of them, only an inflammation of the synovial membrane on the inner side, with no structural damage.
The movement analysis showed no technical differences between those who had pain and those who did not: the authors' conclusion is that it is an overload syndrome, that is, one of quantity.
This changes the message in two directions. On one side it puts the alarm into proportion: breaststroke has not been shown to wear out the meniscus or damage the ligaments in people who swim to stay well. The documented picture is an irritation from repetition in athletes who do thousands of kicks a week. If you do two lengths of breaststroke on a Sunday, you are not destroying anything.
On the other side it tells you how to behave: since it is a problem of dose, the dose is the lever.
If you already have pain on the inner side of the knee, as in the pictures described in pain on the inner knee, it would be wise to keep breaststroke out of the programme in the early phase, or to limit it to a few isolated lengths, building the volume with crawl and backstroke.
The simplest rule there is also applies: if a movement in the water triggers a precise, localised pain, that movement is not for you today, whatever the article says.
One last note that almost everyone forgets: in competitive swimming knee pain is the second most common complaint, and the first is the shoulder. Moving all the load from the legs to the arms does not eliminate the problem, it shifts it. Those who quickly increase lengths to protect the knee often reach the physiotherapist with an inflamed shoulder.
How to combine the pool with strengthening on land
In clinical practice the simplest way to use water well is to treat it as one of the two legs of the programme, not as the programme itself.
- Two water sessions and two land sessions a week is a schedule that holds up in real life. If you can only manage three in total, keep at least one on land.
- In the water put the volume: crawl or backstroke lengths, walking up and down, knee flexion and extension against the resistance of the water. It is the place to move a lot without paying for it.
- On land put the strength and the load: getting up and sitting down from a chair, going up and down a low step, controlled mini-squats, exercises for the quadriceps and glutes with a band or progressive load.
- An important detail: the 2024 Cochrane update found no differences between the various types of exercise, nor a link between results and the total number of prescribed sessions. The type of exercise matters less than the consistency with which you do it.
- Progression has an order: first you increase the quantity at unchanged difficulty, then you increase the difficulty. Changing both at once is the most common way to hurt yourself.
- The move from water to land is gradual, not a switch: you shift one session at a time, keeping an eye on how the knee responds over the following 24 hours.
Swimming and the knee with osteoarthritis: what to expect
If you have osteoarthritis, the pool is a good ally provided you know what it promises. The documented effects are slightly lower pain and slightly better function immediately after the course; three months later, in the most recent meta-analysis, the only effect still present was the one on pain, while on stiffness and function the advantage was not shown to persist.
On quantity, the only serious reference for swimming in this population is the protocol of the randomised trial mentioned above: 45 minutes, three times a week, for twelve weeks, at moderate intensity. It is a good starting point to build from, with the caveat that it was a supervised programme and that you are not the average of 48 people.
On weight, a useful clarification: it is often read that swimming helps you lose weight and therefore unloads the knee. A meta-analysis of twenty studies conducted in healthy adults found that exercise in water improves body composition only at moderate or moderate-to-high intensity and with more than 120 minutes a week; below that threshold the effects were not seen.
That figure has not been collected in people with osteoarthritis, but the order of magnitude is clear: two slow half-hour swims are not a weight-loss strategy.
It is no coincidence that the guidelines for knee osteoarthritis, alongside land-based exercise, allow for combining it with weight management through diet, which in the ACR recommendations is a strong recommendation for people who are overweight.
Swimming and the knee after surgery: the timing is not set by the pool
After an arthroscopy, a reconstruction or a prosthesis, the moment you can return to the water does not depend on a general rule, but on two concrete things.
The first is wound healing: until it is closed and the surgeon gives the go-ahead, the pool remains off limits because of the risk of infection. The second is your surgeon's protocol, which takes into account the repaired tissue and the movements to protect in the first weeks.
Anyone who gives you a number of weeks without knowing what was done to you is guessing.
As for what water adds after a prosthesis, the evidence is honestly weak: one review found only three small studies, with improvements in range of motion and swelling but a functional gain that the authors themselves did not consider clinically relevant. Put practically: water is a welcome complement to land-based rehabilitation, not a substitute.
If swimming is not for you: the alternatives in water and out of it
Swimming requires a technique not everyone has, and swimming badly tires the shoulders without training the legs. The alternatives in the pool work very well and are, in effect, what the studies have tested most.
- Walking in water, forwards and backwards, adjusting the depth as discussed above.
- Knee flexion and extension against the resistance of the water, holding on to the edge.
- Cycling in water, supported by a float board or the pool edge.
- Hydrokinesitherapy proper, that is, a programme built around you and your knee.
And outside the water, remembering that in the randomised trial it gave the same results as swimming: cycling, which has the advantage of costing less and being always available. If the knee complains on the saddle, it is almost always a question of adjustment: a saddle set too low and gears that are too hard are the most common causes.
How to tell if you are getting the right dose
To tell whether swimming is doing your knee good, you do not need percentages. You need two questions, every week.
- Does the discomfort you feel after the session settle by the following day? If yes, the dose is sustainable. If every time it leaves your knee worse for two or three days, you are asking too much: reduce the lengths or change stroke, before giving up altogether.
- Is the trend over three or four weeks improving? Not the single session: the direction. If the curve is flat, the problem will not be solved by swimming more, and the starting hypothesis needs revisiting.
If after three months of regular swimming your knee is unchanged, it is almost never because you are not swimming enough. Much more often what is missing is what the water does not give: strength work and the gradual re-accustoming to load.
In summary
The relationship between swimming and the knee is positive and worth exploiting, but it should be described for what it is. Water is an excellent way to keep moving when land-based load hurts, with real but small benefits on pain and function, equivalent to, and not superior to, those of a good land-based programme or cycling.
Water removes less weight than is claimed if you are not immersed deeply, breaststroke is a problem of quantity rather than a poison, and no stroke replaces the strength work the knee needs to cope with stairs. "The ideal sport for the joints" is a headline, not a research result.
If you have doubts about which stroke to choose, how much to swim, or what to add to the pool so that you do not stay stuck where you are, the physiotherapists at Ri-Hub can assess your case and build the programme with you, entirely online.
Sources
- Aquatic exercise for the treatment of knee and hip osteoarthritis, Cochrane Database of Systematic Reviews, 2016
- Efficacy and safety of aquatic exercise in knee osteoarthritis: a systematic review and meta-analysis of randomized controlled trials, Clinical Rehabilitation, 2023
- Exercise for osteoarthritis of the knee, Cochrane Database of Systematic Reviews, 2024
- OARSI guidelines for the non-surgical management of knee, hip, and polyarticular osteoarthritis, Osteoarthritis and Cartilage, 2019
- 2019 American College of Rheumatology/Arthritis Foundation Guideline for the Management of Osteoarthritis of the Hand, Hip, and Knee, Arthritis and Rheumatology, 2020
- Improved function and reduced pain after swimming and cycling training in patients with osteoarthritis, The Journal of Rheumatology, 2016
- Impact of exercise on articular cartilage in people at risk of, or with established, knee osteoarthritis: a systematic review of randomised controlled trials, British Journal of Sports Medicine, 2019
- Variation in percentage weight bearing with changes in standing posture during water immersion: implication for clinical practice, BMC Musculoskeletal Disorders, 2014
- Breaststroke swimmer's knee. A biomechanical and arthroscopic study, The American Journal of Sports Medicine, 1980
- Knee pain in competitive swimming, Clinics in Sports Medicine, 1999
- The effects of water-based exercise on body composition: a systematic review and meta-analysis, Complementary Therapies in Clinical Practice, 2023
- Effects of aquatic therapy and land-based therapy versus land-based therapy alone on range of motion, edema, and function after hip or knee replacement: a systematic review and meta-analysis, Physiotherapy Canada, 2015
Last source check: August 2026.
Wondering how much it would cost to be followed? The cost depends on how many sessions are needed, not on the single rate: you will find the figures in How much physiotherapy for the knee costs.
Need a consultation?
Start with the questionnaire on Ri-Hub: a few questions about your case, you register, and you see who will follow you: a physiotherapist registered with the Albo, whom you write to in chat whenever you like. If you prefer to talk it through by voice, the first consultation is free, lasts 20 minutes and commits you to nothing: they arrive having already read what you wrote, and together you decide whether and how to go on.
No credit card is needed: the first consultation does not commit you to anything.
