Training

Squats and knee pain: should you stop?

By Luigi · 4 min read

"I can't squat, I've got bad knees." I hear some version of that most weeks, usually from someone who was told years ago that squats wreck your knees and has avoided them ever since. If you get squats and knee pain together, it's a fair worry. But for most people the research points the other way: the right kind of squatting is part of the fix, not the cause.

First, the important bit. I'm a personal trainer, not a physio or a doctor, and this post isn't a diagnosis. If your knee is swollen, locks or gives way, or the pain started with a fall, a twist or a pop, get it looked at by a GP or physio before you load it. Everything below is about the much more common situation: a knee that grumbles when you squat, climb stairs or sit for a long time.

"Deep squats are bad for your knees" — are they?

This is the myth that started most people's fear, so let's deal with it first. A large review of the biomechanics research looked at whether half and quarter squats are actually safer than deep ones^[1]. It found that pressure behind the kneecap peaks at around 90 degrees of bend, and that going deeper spreads the load over a larger contact area rather than piling more onto it. The authors concluded that worries about deep squats causing cartilage damage or arthritis were unfounded, provided the technique is learned properly and the load builds up gradually.

They added one twist I like: people who only ever half-squat often end up using much heavier weights to make it hard, and that's arguably the worse deal for your joints. This is a narrative review rather than a meta-analysis, and much of it is lab and cadaver work, so it tells you about loads rather than about real-world injury rates. But it removes the idea that deep squatting is risky in itself.

If you have knee arthritis, exercise is the treatment

Knee osteoarthritis is the big one in the over-40s, and the evidence here is unusually strong. The 2024 Cochrane review pooled 139 randomised trials with more than 12,000 people^[2]. Compared with no exercise or usual care, exercise probably improved physical function and may have improved pain, and the authors note that every international guideline recommends it. They also found no clear difference between types of exercise. In other words, the best programme is the one you'll actually keep doing.

Two honest caveats. Most of the trials had an unclear or high risk of bias, mainly because you can't hide from someone that they're exercising. And there were slightly more reported adverse events in the exercise groups. That's why I build up gradually rather than throwing anyone in.

Strength seems to matter. A separate meta-regression of 45 trials found that programmes meeting proper strength-training standards built more thigh strength than gentler exercise^[3]. The pattern suggested you'd need roughly a 30–40% gain in quadriceps strength before pain and function improved noticeably. That's an estimate across trials, not a promise for any one person. But it's a strong hint that a few easy leg raises won't get you there; progressive loading probably will.

Pain at the front of the knee

The other common one is patellofemoral pain: an ache around or behind the kneecap, often worse on stairs, squatting or after sitting for a long time. An international consensus statement from patellofemoral pain researchers recommended exercise therapy as the main treatment, especially combining hip and knee strengthening^[4]. Foot orthoses and combined treatments also got the nod for some people; mobilisations on their own and electrotherapy didn't.

That hip part is the bit people don't expect. A lot of front-of-knee pain gets better when we strengthen the glutes as well as the thighs, which is why your programme won't just be squats.

Does it have to be completely pain-free?

Not necessarily, which surprises people. A systematic review of trials in chronic musculoskeletal pain compared exercise where some pain was allowed with exercise kept pain-free^[5]. The painful-exercise groups did slightly better on pain in the short term, and there was no clear difference after that. It was a small evidence base — seven trials and 385 people, covering long-term muscle and joint pain in general rather than knees specifically — so I wouldn't push it further than this: a bit of discomfort during exercise isn't automatically a sign of damage.

The practical rule I use: mild discomfort that settles within a day is usually fine. Pain that's sharp, getting worse session by session, or still there the next morning means we change something.

What I actually do with clients who have sore knees

• Find a depth that's comfortable, then build from there. That might be a box squat to a high bench. The depth comes back gradually as the knee tolerates more.
• Strengthen the hips as well as the thighs. Glute bridges, side-lying and banded work, step-ups and split squats all count.
• Change the squat, don't ban it. Goblet squats, box squats and wider or narrower stances load the knee differently. There's almost always a version that works.
• Warm up properly. Stiff knees often loosen after a few minutes. I've written about which warm-ups actually help.
• Track it, and refer when needed. If something isn't improving over a few weeks, or doesn't feel like ordinary knee pain, I'll tell you to see a physio rather than guess.

If you're over 50 and worried your knees are just "wearing out", it's also worth reading why muscle is worth fighting for after 50. Stronger legs protect far more than your knees.