The Millimetre Question: How Much Does Saddle Height Really Matter?

Rather than guessing, measure your actual position with your own camera.

€5 for one fit, or 10 fits for €25. See pricing →

The number that should change how you think about this

A 2011 review of saddle-height research (Bini et al.) found something most riders find hard to believe: a 5% change in saddle height produces up to a 35% change in knee kinematics.

For a rider with a 700 mm saddle height, 5% is 35 mm. But the relationship is not linear and it does not need to be, the point is that the joint is far more sensitive to saddle height than the rider's own perception is. Your knee notices changes your body reports as "fine".

Across studies, roughly 15% to 50% of cyclists report knee pain depending on the group studied (Bini & Flores Bini, 2018), and in bike-fit practice saddle height is the first variable to check, before cleats and before the front end.

What the newer evidence adds. A 2022 systematic review found strong evidence that saddle height should be set from the knee angle measured while pedalling, but only moderate evidence that changes smaller than about 4% of leg length (roughly 30 mm for most adults) change joint loads more than trivially, and limited evidence on injury risk (Bini & Priego-Quesada, 2022). The millimetre amounts below are bike-fit experience, not measured thresholds.

Why small errors become big problems

The mechanism is repetition. At 85 rpm for two hours you turn the cranks roughly 10,000 times. Every one of those revolutions runs your knee through the same arc.

A position error does not hurt because any single pedal stroke is damaging. It hurts because a slightly wrong angle, repeated ten thousand times, accumulates load in a tissue that had no problem with the first hundred. That is why complaints appear after a predictable number of minutes, and why they arrive when you increase your riding volume rather than when you change your position.

It is also why "it feels fine" is a poor test. Your body reports discomfort, not joint angles, and it reports it late.

What each amount actually does

2 mm. Below the threshold of most riders' perception, and within the noise of how you sit from day to day. Not worth chasing, but worth knowing that it exists, because it explains why one ride feels different from another with no change at all.

5 mm. The smallest change that reliably produces a different sensation. This is the standard adjustment step for exactly that reason: large enough to feel, small enough that your body accepts it. A saddle 5 mm too high is enough to start hip rocking in a rider who is not very flexible.

10 mm. Clearly felt by most riders and enough to cause a complaint over a few weeks. This is the typical size of the error that walks into a bike fit: not dramatic, entirely invisible to the eye, and the difference between comfortable and a knee that aches after an hour.

15 mm and up. Usually obvious in the way you pedal, visible hip rocking if too high, knees coming up high if too low, but riders still tolerate it for months, because the body adapts to bad positions surprisingly well before it complains.

The eye cannot do this

Here is the practical problem. You can look at a rider and see that their leg is roughly straight at the bottom of the stroke. You cannot see the difference between 138 and 146 degrees, and those eight degrees can decide whether a position feels comfortable or leads to front-of-knee pain.

Nor can you see it in a mirror, or judge it from a photograph taken at the wrong moment in the pedal stroke. The angle you need is the angle at the exact bottom of the stroke, and it changes continuously.

This is not a claim that riders are bad at judging things. It is a claim about what eyes can resolve, and eight degrees at a distance is below it.

Why guessing costs more than measuring

Working by feel means changing something, riding for two weeks, and evaluating a sensation. If it did not help, you change something else. Each cycle costs weeks, and you cannot isolate which change mattered because your fitness, your volume and the weather all changed too.

Measuring collapses that to a number. Not because the number is sacred, the target is a range, not a point, but because it tells you which direction to move and how far, on the first attempt rather than the fourth.

What to take from this

Three things worth remembering. Adjust in 5 mm steps, because that is the smallest change your body reliably registers. Do not trust "it feels fine" as evidence that a position is correct, because your body reports late. And write down your numbers, because next season you will not remember them and the drift is invisible.

A BikeFit/LAB analysis measures your knee extension, knee flexion, hip angle, torso lean, elbow angle and saddle fore/aft from a short video, and returns the adjustments in millimetres against the target range for your riding goal. The value is not the report, it is skipping the four cycles of guessing.

Rather than guessing, measure your actual position with your own camera.

€5 for one fit, or 10 fits for €25. See pricing →