The Ache That Shows Up At Kilometre Forty
You feel strong for the first hour. Somewhere past Papan on the way back into Ipoh, a dull ache settles in behind the kneecap. It is not sharp enough to stop you, so you ride home on it. By evening it has stiffened up, and the Tuesday ride does not happen.
Knee pain is the complaint bike fitters hear more than any other, and unlike most cycling niggles it has a habit of taking your riding away from you rather than just annoying you.
In a study of 109 professional road cyclists interviewed about the previous twelve months, 36% had experienced anterior knee pain and 19% had sought medical attention for it. More telling than the prevalence: of all the injuries that actually caused time off the bike, 57% were knee injuries. Lower back pain was more common overall, but the knee is what stops you riding.
Across the wider population of riders the figure is higher still. A 2011 review of the saddle height literature put the occurrence of overuse knee joint pain in cyclists at around 50%.
The good news is that most of it is a setup problem. The less convenient news is that the research is messier than most bike fit marketing admits, so this post separates what is well evidenced from what is trade convention.
What Cyclists With Knee Pain Actually Look Like On A Bike
This is the finding that surprises people.
A systematic review looked at every original study comparing cyclists with knee pain against cyclists without it. The pattern that came out was not the one you would expect:
- Riders with knee pain showed greater knee adduction and an increased medial projection of the knee. In plain terms, the knee drops inward toward the top tube through the pedal stroke.
- They showed greater ankle dorsiflexion and altered timing between the vastus medialis and vastus lateralis, the two quadriceps muscles that steer the kneecap.
- There was no difference in knee flexion angle between the painful and pain-free groups, and no clear difference in patellofemoral or tibiofemoral joint forces.
Read that last point again, because it undercuts a lot of received wisdom. The single number most fitters quote at you, how bent your knee is at the bottom of the stroke, did not distinguish riders in pain from riders who were fine.
What did distinguish them was the path the knee travels, viewed from the front. Not how far it bends, but whether it tracks straight or falls in.
The honest caveat: these are retrospective comparisons. They tell you what a painful knee looks like, not that the tracking caused the pain rather than the pain causing a compensation. That distinction matters and it is why nobody should promise you a cure.
Saddle Height: Real, But Not The Universal Answer
Saddle height is where the strongest numbers live, and where the most overclaiming happens.
The 2011 review found that across the available studies, a 5% change in saddle height altered knee joint kinematics by around 35% and knee joint moments by around 16%. It also found that patellofemoral compressive force is inversely related to saddle height: drop the saddle and the pressure between kneecap and femur goes up. That is a clean mechanical explanation for why a low saddle produces pain at the front of the knee.
Its recommendation was to set saddle height by dynamic knee flexion angle, in the region of 25 to 30 degrees at the bottom of the pedal stroke, rather than by any of the inseam formulas.
Then the same lead author ran an updated systematic review in 2022, screening nearly 30,000 papers and including 41. Two conclusions worth having:
- Strong evidence supports setting saddle height from dynamic measurement of the knee angle, and limb kinematics genuinely do change with saddle height.
- Moderate evidence says that changes smaller than 4% of leg length produce trivial to small changes in lower limb loading, with no effect on oxygen uptake or efficiency.
That second point is the useful one for riders. For a typical rider, 4% of leg length is somewhere around 3 centimetres. So the “5mm changed my life” stories you read on forums are not what the loading data supports. If your saddle is out by 5mm, that is worth correcting for comfort and consistency, but it is unlikely to be the thing generating your knee pain. If it is out by 2 or 3 centimetres, which is common on bikes that were never set up at all, that is a different conversation.
The review was also blunt that direct evidence on saddle height and injury risk specifically remains limited. We have good kinematic data and weak outcome data.
Where It Hurts, And What Fitters Read Into It
The pain-location map below is standard practice among fitters and clinicians. It is clinical reasoning built on the mechanics rather than something proven in a trial, so treat it as a starting hypothesis, not a diagnosis.
Front of the knee, around or under the kneecap. The most common presentation. Associated with a saddle that is too low or too far forward, both of which increase patellofemoral compression. Often worse on climbs and into a headwind, where you are grinding a big gear at low cadence.
Behind the knee. Associated with a saddle too high or too far back, over-extending the leg and loading the hamstring attachments. Often accompanied by hips rocking side to side on the saddle, which is the tell your fitter looks for.
Outside of the knee. Classically iliotibial band related. Associated with cleat rotation forcing the foot into an unnatural angle, and with excessive stance width.
Inside of the knee. Associated with the knee collapsing inward under load, which links back to the medial projection pattern the systematic review actually found. Cleat lateral position and stance width are the usual suspects.
If your pain is one-sided, that is information. Bodies are asymmetric, bikes are not, and a fit that treats both legs identically will load an asymmetric rider unevenly. We covered this in 5 Hidden Bike Fit Mistakes New Cyclists Make.
Cleats Change Your Knee Angle Without Changing Your Power
Here is a study that reframes what cleat position is for.
Twelve competitive road cyclists rode with the cleat at the first metatarsal head, then 15mm forward, then 15mm back, at three intensities plus a sprint. The result: no significant change in oxygen cost, pedalling kinetics, muscle activation across six muscles, sprint performance or perceived exertion.
But hip, knee and ankle angles all changed. Knee extension angle in particular was altered by the anterior cleat position.
So the trade-off riders assume they are making, sliding cleats around to chase watts, is not the trade-off in front of them. Fore-aft cleat position is a joint angle lever, not a power lever. You are choosing how your knee moves, at no measurable cost to performance. That is a good deal if the knee is what hurts.
Crank Length: The 2026 Trend With Evidence Behind It
Short cranks are the fit conversation of the year, and unusually for a trend, there is fresh data.
Fourteen amateur cyclists were tested at five crank lengths: their preferred length, plus and minus 5mm, and plus and minus 10mm, with full three-dimensional motion capture. Findings:
- No effect on mean power output across any crank length.
- Shorter cranks produced significant decreases in knee flexion and hip flexion.
- Shorter cranks produced more anterior pelvic tilt and less pelvic obliquity, hip abduction and knee rotation.
The authors concluded that minor reductions in crank length can reduce hip and knee flexion and limit out-of-plane motion, and could serve as an effective way to reduce overuse injury risk.
Two things follow. If you have limited hip mobility and your knee comes up very high at the top of the stroke, shorter cranks reduce that range for free. And the reduction in knee rotation speaks directly to the tracking problem the knee pain research identified. It is not a cure-all, and 5 to 10mm is a modest change, but it costs you nothing in power.
Does A Bike Fit Actually Fix It?
The most relevant study is Canadian. Seventy-six recreational road cyclists who came to a physiotherapy clinic specifically for cycling knee pain received a comprehensive fit focused on knee flexion and knee alignment relative to the pedal axis, plus education on cadence and training progression, plus prescribed exercises.
At twelve weeks, on a 0 to 10 pain scale:
- Worst knee pain improved by 2.52 points
- Average knee pain improved by 1.81 points
- Least knee pain improved by 0.70 points
All statistically significant. We wrote about this and the wider evidence base in What Research Says About Bike Fitting and Injury Prevention.
The caveat you should hear from us rather than find out later: this was a single group observational study with no control arm. Everyone got the intervention, so some of that improvement could be natural recovery or regression to the mean, and the intervention bundled fitting with exercise and education. It is genuinely encouraging, and it is the best evidence we have for fitting and knee pain, but it is not a randomised trial.
What To Try Before You Book Anything
In order of likely impact for the effort:
- Check your saddle height is not badly wrong. Heel on the pedal at the bottom, leg straight with no hip rocking, then clip in normally. If you are hips-rocking or heels-dropping to reach, you are out by enough to matter.
- Drop a gear and spin. Grinding at 60rpm loads the knee far harder than spinning at 85rpm for the same speed. Cadence was part of the intervention in the study above for a reason. This is free and it works fastest.
- Look at your knees from the front. Film 20 seconds on a turbo or have someone watch you. If a knee swings in toward the top tube, that is the pattern the research associates with pain.
- Loosen up your cleat rotation. If you are on zero-float cleats and your knee hurts on the outside, that is a cheap thing to change.
- Back off the volume for two weeks. Overuse means what it says. Big jumps in weekly distance are a trigger regardless of setup.
If the pain persists after two or three weeks of doing all five, the cause is structural to your position and guessing further tends to relocate the problem rather than fix it. That is the argument for measuring it, which we made in Professional Bike Fit vs YouTube DIY.
When To See A Doctor First
A fitter is the right call for pain that appears during or after riding and settles with rest. See a doctor before a fitter if you have:
- Swelling, locking or the knee giving way
- Pain that is present walking or on stairs, not just cycling
- A sudden onset with a specific moment of injury, rather than gradual build-up
- Pain that does not settle at all with two weeks of reduced load
Bike fit addresses load distribution. It does not address a meniscus tear or an inflammatory joint problem, and a good fitter will tell you that rather than sell you a session.
The Short Version
The evidence supports three things clearly. Cyclists with knee pain track their knees differently, falling inward under load, rather than simply bending them more or less. Saddle height matters at the scale of centimetres rather than millimetres, and should be set dynamically. Cleat position and crank length change how your knee moves at no measurable cost to your power, which makes them the cheapest levers you have.
It does not yet support anyone promising that a fit will cure your knee. What we have is one good observational study showing meaningful improvement, a solid mechanical rationale, and a lot of riders who stopped hurting. That is worth being straight about.
For an overview of the other common cycling complaints and what causes them, see our guide to cycling pain.
References
- Clarsen, B., Krosshaug, T., & Bahr, R. (2010). Overuse injuries in professional road cyclists. American Journal of Sports Medicine, 38(12), 2494-2501. PubMed: 20847225. The 109-rider prevalence and time-loss figures.
- Bini, R. R., & Flores Bini, A. (2018). Potential factors associated with knee pain in cyclists: a systematic review. Open Access Journal of Sports Medicine, 9, 99-106. PMC5973630. Knee adduction, medial projection, and the absence of a knee flexion angle difference.
- Bini, R., Hume, P. A., & Croft, J. L. (2011). Effects of bicycle saddle height on knee injury risk and cycling performance. Sports Medicine, 41(6), 463-476. PubMed: 21615188. The 5% / 35% / 16% figures and the 25 to 30 degree recommendation.
- Bini, R., & Priego-Quesada, J. (2022). Methods to determine saddle height in cycling and implications of changes in saddle height in performance and injury risk: a systematic review. Journal of Sports Sciences, 40(4), 386-400. PubMed: 34706617. The 4% of leg length threshold.
- Chartogne, M., Millour, G., García-López, J., Duc, S., Rodríguez-Marroyo, J. A., Pernía, R., & Bertucci, W. (2023). Acute effects of small changes in antero-posterior shoe-cleat position on physiological and biomechanical variables in road cycling. Sports Biomechanics, 22(4), 510-521. PubMed: 35129429. Cleat fore-aft changes joint angles but not power.
- Reynolds, S., Chidley, J., Briley, S., & Outram, T. (2026). The impact of minor crank length adjustments on lower body cycling kinematics. Sports Biomechanics, 25(3), 454-467. PubMed: 40464620. Crank length, knee flexion and pelvic tilt.
- Viau, A., Tremblay, C., Coutu, G., Desmeules, F., & Lafrance, S. (2025). Impact of physiotherapy-led bike fitting on the evolution of knee pain in recreational cyclists: the PBF study. Physiotherapy Theory and Practice, 41(9), 1761-1770. PubMed: 39982032. The 76-rider fitting outcome study.
Ready to Sort It Out Properly?
If you have tried the five checks above and the ache still turns up at kilometre forty, the next step is measuring how your knee actually tracks rather than guessing at it. Chat with our AI bike fit expert on WhatsApp for a free assessment. Tell us exactly where the pain sits, when in the ride it starts, and what you have already changed.