Running builds the strongest aerobic engine on earth. It also loads your knees at two to three times your bodyweight, around 5,400 times every 30 minutes.
Both of those things are true at once, which is why "is cycling better for my knees than running?" deserves a proper answer instead of the two lazy ones the internet offers — the cycling forum's of course, running destroys you and the running forum's nonsense, running strengthens joints. Each is half right. The interesting part is which half, and what it means for your training if you're the kind of athlete whose knees have started sending memos.
Let's take the evidence in order: the mechanics, the arthritis data, the ways cycling can still hurt a knee, and the one argument running wins outright.
The impact arithmetic
Start with what each sport physically does to the joint.
Every running stride ends with a collision. Your foot decelerates from forward motion to a dead stop in milliseconds, and the ground pushes back with a force of two to three times your bodyweight. That force travels up through the ankle, the tibia, and into the knee, 80 to 100 times per minute per leg. Run for half an hour at a typical cadence around 180 steps per minute and your lower body absorbs roughly 5,400 of those impact events. A 75kg runner is processing something in the region of 150-225kg of peak force, thousands of times, every single session.
Your body is genuinely brilliant at this. Tendons store and return energy, cartilage distributes pressure, bone remodels itself stronger along the stress lines. Impact is a stimulus as well as a cost — hold that thought, because it matters later.
Now the bike. Cycling is non-weight-bearing: the saddle carries your mass, and the knee never experiences an impact peak because there is no collision anywhere in the pedal stroke. Load is applied smoothly, through a fixed plane of motion, rising and falling with each revolution like a wave rather than a hammer. The knee flexes and extends through a controlled, repeatable arc, under muscular force you choose with your gearing.
Same joint, two completely different mechanical lives. One absorbs thousands of impact spikes per session; the other experiences continuous, moderate, shock-free load. Whatever else the studies say, that's the physics underneath them.
What the arthritis evidence actually shows
The headline number comes from research highlighted by Stamford Health's orthopedics group: cyclists are 21% less likely to develop knee arthritis than the general population. Cycling moves the knee through repeated, moderate load without impact — and cartilage, which has no blood supply and feeds itself through compression-and-release cycles, appears to like exactly that. Movement without violence.
But before this becomes an obituary for running, the running data deserves a fair hearing — because it's better than most cyclists think.
The largest analysis on the question, a 2017 meta-analysis by Alentorn-Geli and colleagues covering more than 114,000 people, found that recreational runners had lower rates of hip and knee osteoarthritis — around 3.5% — than sedentary non-runners at 10.2%. Sitting on the couch was roughly three times worse for the joints than running moderate mileage. Moderate impact seems to condition cartilage and strengthen the structures around the joint rather than grind them down.
The same analysis found the risk climbing again at the far end: high-volume, competitive running for years pushed arthritis rates above 13%. The dose-response curve is a U-shape. None of the load is bad for joints. Moderate load is protective. Extreme load, sustained for long enough, starts collecting.
Put the two findings together and the picture sharpens. Healthy knees tolerate — and probably benefit from — sensible running volumes for decades. Some runners genuinely do run injury-free into their seventies, and nothing in this article argues they should stop. The evidence shifts decisively toward cycling in three situations: when volume is high, when the knee has previous damage (a meniscus tear, an old ACL, early cartilage wear), or when pain keeps returning despite sensible training. In those cases, the 21% figure isn't a curiosity — it's a strategy.
Where cycling still stresses the knee
Now the part the cycling forums skip. Low impact is not zero stress, and any cyclist who has ground up a climb in too big a gear knows precisely where the knee feels it.
The mechanism is patellofemoral compression. Every pedal stroke, your quadriceps fire to drive the pedal down, and that contraction presses the kneecap into its groove on the femur. The force is smooth — no impact spike — but it's real, and two things multiply it.
The first is high resistance at low cadence. Grinding a huge gear at 55 rpm means enormous quadriceps force per stroke, and the patellofemoral joint takes that compression every revolution. This is why new cyclists who muscle everything in big gears develop front-of-knee pain, and it's why torque work — a genuinely useful training tool that coaches like Dan Lorang and John Wakefield prescribe to World Tour riders — belongs in a conditioned cyclist's programme, not a beginner's first month. Cadence in the 85-95 rpm range keeps per-stroke forces modest and shifts the work onto the cardiovascular system, which is far better equipped to absorb it.
The second is bike fit — above all, saddle height. A saddle set too low increases the knee flexion angle through the whole stroke, and deeper flexion under load means higher patellofemoral compression, thousands of times per hour. It's the classic self-inflicted cycling knee injury, and the fix costs nothing but attention. (A saddle too high creates its own problems at the back of the knee.) This is why a professional bike fit is the single best knee-protection purchase in cycling, and why our guide to what to check first when your knee hurts on the bike starts with the seatpost, not the physio.
Set up properly, ridden at sensible cadences, cycling is about as knee-friendly as meaningful exercise gets. But it earns that status through mechanics, not magic — get the mechanics wrong and the bike will remind you it's still a repetitive-load sport, roughly 5,000 knee flexions per hour.
The argument running wins
There's one place where this comparison flips completely, and it deserves a full section rather than a footnote: your skeleton.
Bone is like muscle — it adapts to load, and it wastes without it. Running's impact, the very thing your knees resent, is exactly the stimulus bone needs to maintain density. Cycling provides almost none of it. You're seated, supported, and shock-free, which is wonderful for cartilage and useless for bone.
The data on this is genuinely uncomfortable reading for lifelong cyclists: masters cyclists are seven times more likely to have osteopenia of the spine than runners of the same age. Cycling exclusively for years, especially through the decades when bone density naturally declines, leaves the skeleton quietly under-built — and nobody notices until a routine crash breaks something that shouldn't have broken. We've covered cycling's bone density problem and how running fixes it in full, but the summary is simple: the bike protects your knees at the price of neglecting your bones, and running pays the bone bill better than almost anything.
Which collapses the whole running-versus-cycling debate into a much better idea: don't pick.
The evidence-based combination for joint and skeletal health is both sports, weighted by what your body currently needs. Cycling carries the bulk of the aerobic volume — the engine work, the long endurance hours, the intervals — with no impact cost. Running survives as a small, protected dose: one or two short, easy sessions a week, enough to keep loading bone and maintaining impact tolerance, nowhere near enough to reignite the knee trouble that started this conversation. A dose you could hold for another thirty years.
Protecting your knees in practice — on the bike and off it
The evidence above converts into a short, unglamorous checklist. None of it is expensive, and all of it beats managing pain after the fact.
Set the saddle height properly. The rough field method: sit on the bike with your heel on the pedal at the bottom of the stroke — your leg should be dead straight. When you clip in normally, that leaves a bend of roughly 25-35 degrees at the knee at full extension, which is the range most fitters target. If your hips rock across the saddle, you're too high; if your knee stays deeply bent at the bottom, you're too low — and too low is the setting that quietly loads the patellofemoral joint all ride. A professional fit ($150-300) does this with far more precision and checks cleat position and saddle setback while they're at it. For a joint you can't replace cheaply, it's the best money in the sport.
Ride at 85-95 rpm by default. High cadence keeps per-stroke quadriceps force low and moves the work to your heart and lungs. Save deliberate big-gear torque efforts for when you're conditioned, and build into them gradually — they're a tool with a real payoff and a real patellofemoral cost.
Progress load like an adult. Cartilage and tendon adapt slower than fitness. When you add volume or intensity, add it in modest weekly steps and let the tissue vote before you add more.
Do strength work. This is the piece most endurance athletes skip, and it protects the knee in both sports. Stronger quadriceps and glutes control the kneecap's tracking and absorb load the joint would otherwise take raw; hip strength in particular keeps the knee from collapsing inward under fatigue, on the run and on the pedal stroke alike. Two short sessions a week of squats, step-ups, lunges and hip work covers it — the complete strength guide for cyclists lays out the progression. Nothing in this article, cycling included, protects a knee as reliably as the muscle around it.
And structure the two sports so they don't compete. If you're keeping the hybrid week — most volume on the bike, a small preserved running dose — put the runs on days after easy rides rather than hard ones, and never stack a run onto legs still wrecked from intervals. A combined running and cycling schedule is easy to get right and easy to get wrong, and the difference is mostly sequencing.
For the rider between 35 and 55
If you're in the audience this article is really for — the athlete in their 40s whose knees have started negotiating — one more piece of context matters.
Knee cartilage has essentially no blood supply and very little capacity to regenerate. The cartilage you have now is, for practical purposes, the cartilage you're finishing the ride with. That's not a reason for fear — cartilage responds well to moderate cyclic loading, which is precisely what both sensible running and well-set-up cycling provide. But it does change the cost of ignoring repeated warnings. A knee that hurts after most runs, settles, and hurts again isn't asking for tougher mental framing. It's reporting load it can't currently absorb.
Moving your aerobic volume onto the bike in response isn't surrender, and it isn't the end of anything. It's the same decision a good team manager makes with an engine worth protecting: change how the miles get done. Your VO2max doesn't care which sport maintains it. Your discipline, your zones, your capacity to suffer usefully — all of it transfers to the bike and keeps compounding, and there's a strong argument you can get properly fast on the bike after 40 — faster, over the long arc, than knees-permitting running was ever going to allow.
So — is cycling better for your knees than running? Mechanically, yes, and it isn't close. Completely? No: running keeps your skeleton honest, and moderate running never deserved its joint-destroyer reputation in healthy knees.
The strongest answer was never either/or. Keep the engine on the bike. Keep a little impact in the week. And let your knees spend the next few decades doing what they were complaining they couldn't: lasting.