Gravel crank length: what actually changes when you go shorter

Power is force times the speed the pedal travels. Shorten the crank and the same force produces 4.35% less torque, but the pedal also covers less distance per revolution: 1036.7 mm on a 165 mm crank against 1083.8 mm on a 172.5 mm crank. Hold your cadence and foot speed drops; let cadence rise about 4.5% and foot speed, and therefore power, is back where it started.

Everything else follows from the 7.5 mm. The pedal at the bottom of the stroke sits 7.5 mm higher relative to the bottom bracket, so the saddle goes up 7.5 mm to keep the same leg extension. At the top of the stroke those two changes add rather than cancel: the saddle is 7.5 mm higher and the pedal 7.5 mm lower, so there is about 15 mm more distance between hip and pedal, and the knee and hip are less closed at the top.

At 60% of VO2max for 20 minutes, perceived exertion came out at 3.21 on 165 mm against 4.50 on 175 mm (p < 0.001), while oxygen consumption (42.64 vs 43.29 ml/kg/min) and heart rate (146.7 vs 148.7 bpm) were statistically indistinguishable. Sprint power was flat too: 13.23 vs 13.18 W/kg ¹.

Two caveats the summaries usually drop. First, 172.5 mm was never tested — the trial ran 165, 170 and 175 mm, and 165 vs 170 was NOT significant (p = 0.25). The honest reading is that a big jump down from 175 mm changes how the effort feels; a small one from 172.5 mm to 170 mm probably does not. Second, the participants were 28 trained male cyclists, so it says nothing directly about untrained riders or about women.

That is still a useful result, because it kills the assumption shorter cranks cost power. It just does not make 165 mm universally correct, and it is not a medical finding. If you have knee or hip pain, that is a conversation with a bike fitter or a clinician, not a part you can buy your way out of.

Martin and Spirduso tested maximal cycling power across a wide span of crank lengths and concluded that pedalling rate and pedal speed are what determine maximal power, with crank length itself a minor term ². McDaniel and colleagues reached the equivalent conclusion for submaximal riding: metabolic cost tracks mechanical power and pedal speed ³. That is the mechanism behind the cadence compensation — match foot speed and you have matched the demand.

Barratt and colleagues looked underneath the total and found the joint-level contributions do shift with crank length even when the total does not ⁴, which is the measurable version of what riders describe as a different feel rather than a different ceiling. Ferrer-Roca and colleagues tested SMALL changes in crank length and found gross efficiency and pedalling technique barely moved ⁵ — more support for treating 172.5 mm to 170 mm as a non-event and 175 mm to 165 mm as the change worth making.

On the generalisability gap: Macdermid and Edwards ran crank length on well-trained female cross-country mountain bikers ⁶, which is the population the 2025 trial did not cover. Worth reading if you are sizing cranks for a rider the male-only studies do not describe.

The pedal sits 7.5 mm higher at the bottom of the stroke, which is the difference between clearing a rock mid-corner and not. On terrain where you are pedalling through ruts and over roots, that margin gets used constantly.

Toe overlap is the other one. A 165 mm crank puts the pedal 7.5 mm further back at the forward position, which on a small or medium frame running a 45 mm-plus tyre can be the difference between your shoe catching the front tyre in a tight turn and not. Both are geometry, not physiology — they apply whatever the research says about power.