Cycling knee pain: the mistake that cost me a race

It always starts the same way — a faint itch on the inside of my left kneecap, ten minutes into a ride. Not pain yet. Just a signal. I’ve had cycling knee pain enough times to know exactly what that itch becomes if I ignore it: inflammation around the patella, a knee I can’t load, and weeks off the bike. This year I ignored it anyway, and it cost me my target race.

The mistakes are more useful than the diagnosis, so here is the whole sequence.

The itch is the injury talking early

Pain at the front of the knee — around and behind the kneecap — is the most common overuse complaint in cycling. It doesn’t arrive from a single moment. It builds. Which means it also warns you, and the warning is that small sensation most riders ride straight through.

I know that signal. I still didn’t stop. The rule I’ve since made non-negotiable: if I feel the itch around the patella, the ride is over that minute. A skipped session costs a day. Riding through the itch costs a month. I’ve now paid the month often enough to respect the day.

How I dug the hole

Two weeks out from my A race — the local XCM marathon championship — I was training hard and tipped into overtraining. Then a family holiday stacked several days of heavy hiking on top of it. Fatigue with no floor under it. The knee was already primed before I did anything obviously wrong.

Then the stupid part. After cleaning the mountain bike I noticed my saddle-height marker tape had slipped, so I reset it — and put the saddle roughly a centimetre too high. A centimetre sounds like nothing. It isn’t. Saddle height changes how the knee tracks through every pedal stroke, and even a few millimetres in the wrong direction loads the joint differently across thousands of revolutions. On a fresh knee I might have gotten away with it. On an already-irritated one it was fuel on a fire I hadn’t noticed was lit.

I rode out to recon the race course. The itch showed up early — a 2 out of 10. I told myself it would settle. It didn’t. I finished at an 8 out of 10, with a knee I couldn’t load without wincing.

Rest, test, rest — the recovery mistake that actually cost the race

For the next two weeks I did the worst possible thing. Rest a few days, feel better, test it with a ride, reirritate it, repeat. I was afraid of losing form so close to the race, so I kept prodding the knee to check whether I still had one to race on. What I actually did was keep the inflammation alive and never let it fully calm down. The stop-start prolonged the recovery instead of shortening it.

I lost the race. Not by finishing badly — I couldn’t start.

Once a joint is inflamed, “testing it” is not recovery. It’s re-injury on a timer. Full rest followed by a controlled, graded return beats poking at it every three days for reassurance. The form I was so afraid of losing came back in a couple of weeks anyway. The race did not come back at all.

What actually fixed it

Five weeks after that first itch, it was resolved. The fix was not complicated:

Five days of complete rest from the bike — no testing.
Easy spinning only on return, kept well below any load that provoked the knee.
Resistance-band work for the glutes and the outer-quad and hip stabilisers — the muscles that hadn’t been doing their share.
A physio visit with an ultrasound. Nothing structural: no tear, no tendon damage. Just inflammation from the patella being pressed into its groove because the muscles around it were imbalanced.
A personalised program to build the underdeveloped muscles, plus activation and stretching work for the tendons.

That diagnosis is the point of this post. My quads were strong — years of intervals and volume will do that. My glutes were weak. When the glutes don’t do their share, the load transfers to the quads, and the quads pull the kneecap harder into its groove. The strength literature for patellofemoral pain names the weak links almost in order: quads, glute medius, glute max, hamstrings. I had trained none of them directly. Only cycling, only intervals. It bit me.

The 44/56 split I couldn’t see

While rebuilding, I finally bought the second pedal for my power meter. Until then I’d measured only the left side and doubled it. The dual-sided reading was ugly — 44% left, 56% right. My weaker, painful leg was also my lazy leg, and a single-sided meter had given me no way to see it.

Some left/right imbalance is normal, and small differences aren’t worth chasing. Mine wasn’t small. With a live balance number on the head unit and some deliberate single-leg focus, four weeks later it sat between 51/49 and 49/51. Faster than I expected — maybe I got lucky. But you can only correct what you can measure, and the exact asymmetry feeding the injury had been invisible until I paid to see it.

Strength is not optional cross-training

Superfly is built on a simple idea — hard days hard, easy days easy, no grinding in the middle. Strength work is the pillar I’d quietly treated as optional, and the knee is where that debt came due. Intervals and volume make you fitter; they do not make you balanced. The muscles that stabilise the knee and hip need direct, progressive loading, and tendons respond to slow, heavy, controlled work rather than more pedalling. I took that last part from Scott Hogan’s Built from Broken — written for lifters, but the joint and tendon principles map straight onto cycling. Best thing I read the whole time I was hurt.

Two short glute-and-stabiliser sessions a week would have cost me almost nothing. Skipping them cost me a season’s target race.

FAQ

Why does cycling cause pain around the kneecap?

Front-of-knee pain in cyclists is usually patellofemoral — the kneecap being pressed into its groove under repeated load. The common driver is a muscle imbalance: strong quads and weak glutes, so the quads take over and pull the patella harder into the joint. Bike fit and sudden jumps in training load make it worse.

Should I keep riding through cycling knee pain?

No. Front-of-knee pain builds gradually and warns you early with a faint sensation before it becomes real pain. If you feel that itch around the kneecap, end the ride — a skipped session costs a day, riding through it can cost weeks. Testing an already-inflamed knee every few days prolongs recovery rather than shortening it.

Can a saddle 1 cm too high cause knee pain?

Yes. Saddle height changes how the knee tracks through every pedal stroke, and even a few millimetres off can alter the load on the joint across thousands of revolutions. A change that a healthy knee tolerates can push an already-irritated one over the edge. Re-check your saddle position after any maintenance that might have moved it.

What strength exercises help cyclist’s knee?

The priority muscles for patellofemoral pain are the quads, glute medius, glute max, and hamstrings. Resistance-band work for the glutes and hip stabilisers, plus single-leg loading like split squats and bridges, targets the links cycling leaves underdeveloped. Two short sessions a week is enough to shift the balance over a few weeks.

How do I fix a left/right power imbalance on the bike?

First you need a dual-sided power meter — a single-sided unit hides the asymmetry entirely. Put the live balance number on your head unit, add single-leg pedalling drills, and strengthen the weaker side directly. A meaningful imbalance can move back toward 50/50 within a few weeks, though small differences are normal and not worth obsessing over.

How long does patellofemoral knee pain take to recover in cyclists?

It depends on how early you catch it and how disciplined the rest is. Mine took five weeks from the first warning, and most of that was self-inflicted by testing it too often. With genuine rest and a graded return plus strength work, a mild case can settle in one to two weeks.