Tyre width and rolling resistance: why 28 mm beat 23 mm

For about thirty years the received wisdom was that narrow tyres roll faster. It was wrong, and the way it was wrong is instructive: the comparisons that produced it were not comparing width. They were comparing width and pressure and rim and casing all at once, and attributing the result to the only number printed on the sidewall.

Run the test properly — same casing, same pressure, same load, same surface, only the width changed — and the wider tyre comes out with the lower coefficient of rolling resistance. Not by a huge margin, but consistently, and for a reason that falls straight out of the geometry of a flattened tube.

This post is about that mechanism, about where the advantage stops, and about the rim underneath, which matters more than most riders realise. If you want the pressure side of the argument, our piece on tyre pressure covers it and this one deliberately does not repeat it.

Rolling resistance is hysteresis, not friction

The intuitive model — the tyre rubs on the road, rubbing costs energy — is almost entirely wrong for a rolling wheel. A tyre that is not sliding loses very little to friction at the road surface. What it loses energy to is deforming itself.

As each section of tyre rotates into the contact patch it is squashed flat; a moment later it is released and springs back. Rubber and the rubberised fabric of the casing are viscoelastic, which means the energy that comes back out is less than the energy that went in. The shortfall becomes heat. Do that a hundred times a second and you have rolling resistance. Studies that model tyre rolling loss from the material properties upwards, and experimental work on racing bicycle tyres on test rigs, both land in the same place: the loss is dominated by cyclic deformation of the structure.

That framing tells you immediately what to change. If the loss is proportional to how much the casing has to bend, then a tyre that bends less per revolution rolls better. Width is a way of bending less.

Same area, different shape

Here is the piece of arithmetic that the old wisdom missed. A pneumatic tyre supports its load almost entirely by air pressure acting on the flattened contact patch. To a first approximation:

contact patch area ≈ load ÷ inflation pressure

Both sides of that equation are unchanged when you fit a wider tyre at the same pressure with the same rider on top. The contact patch area stays the same. What changes is its shape. A narrow tyre makes a long, thin patch. A wide tyre makes a short, broad one.

The long patch is the expensive one. To make a 43.1 mm flat spot, a 23 mm tyre has to deflect deeply into itself; the casing bends through a sharper arc and more of its volume is worked. The 28 achieves the same supporting area with a flat spot of 33.3 mm and a shallower deflection. Less bending per revolution, less hysteresis, less heat.

Magnitude, honestly: at matched pressure, the improvement from 23 to 28 in controlled testing is usually in the region of five to ten per cent of the rolling resistance coefficient. For an 80 kg system at 30 km/h with Crr around 0.004, that is roughly 1–3 watts. It is real. It is not transformative on its own. What makes width worth having is the second effect.

The loss that drum tests cannot see

A laboratory rolling-resistance test runs the tyre on a smooth steel drum. Real roads are not smooth, and the difference is not a detail.

On a rough surface the tyre can either absorb the texture — casing flexing, wheel path staying level — or it can transmit it, lifting the whole bike and rider vertically. The second costs far more energy, because the mass being moved is you, and the damping is done by soft tissue rather than by rubber. Whole-body vibration measurements on road cycling show just how much energy gets fed into the rider on ordinary tarmac, and modelling work on tyre–road interaction shows how strongly surface texture changes the rolling loss.

This is the loss riders call impedance or suspension loss. It rises steeply with pressure and falls with tyre volume, which is precisely the opposite of how hysteresis behaves. Add the two together and total loss is a U-shape with a minimum at some pressure — and a wider tyre moves that minimum lower and makes the bottom of the curve flatter and more forgiving. That forgiveness is the real prize. A 28 at slightly the wrong pressure is much less punished than a 23 at slightly the wrong pressure.

Where the aerodynamic penalty bites

Width does cost you something, and it is aerodynamic rather than mechanical. Two effects, of very different size.

The first is frontal area, and it is nearly negligible. Going from 23 to 28 mm adds about 5 mm to the width of an object that is perhaps 2% of your total frontal area. Reviews of bicycle aerodynamics consistently put the whole bicycle at roughly a fifth to a quarter of total drag, with the rider making up the rest, and the tyre is a small slice of the bicycle's share.

The second matters more: the junction between tyre and rim. If the tyre is wider than the rim's external width, the airflow that has been following the tyre has to negotiate a step back inwards at the rim shoulder, and it tends to separate there. Wheel makers describe this with a rule of thumb — external rim width should be at least about 105% of the measured tyre width. That rule comes from commercial wind-tunnel testing, not from peer-reviewed work, and you should treat the exact figure accordingly. The underlying phenomenon, flow separation at an abrupt widening, is ordinary fluid mechanics and is not in doubt.

In practice this means the aerodynamic argument against width is really an argument about matching. A 28 on a modern wide-profile rim is aerodynamically fine. A 32 crammed onto a narrow rim from 2012 is not, and it will also handle worse.

The rim underneath

Internal rim width sets how wide your tyre actually sits, how square its profile is, and how far you can drop the pressure before the casing starts folding under cornering load. The effect is measurable: in cross-country mountain biking, testing across different rim widths with the same tyres found rim width changing both rolling resistance and off-road speed.

Two practical consequences. First, measure, do not read. A tyre labelled 28 on a 19 mm internal rim often measures 29–30 mm, and frame clearance is decided by the measured figure plus room for grit. Second, a tyre near the top of its rim's range is being stretched taut and loses some of the compliance you bought it for; a tyre near the bottom of the range balloons and gets vague. The middle of the range is where tyres behave.

So what should you fit?

• Smooth road, racing, 19 mm internal rims: 28 mm. This is where road cycling has settled for good reasons, and the rolling and comfort gains over a 25 outweigh the aerodynamic cost on a rim that matches.
• Ordinary British or European back roads: 30–32 mm. Once the surface is patched, chipped or gravel-strewn, impedance dominates and volume wins comfortably.
• Wet and cold riding: width helps less than casing and compound do. What matters in the rain is the rubber and the pressure, which is covered in riding in the rain.
• Mixed surfaces: go by the worst surface on the route, not the average. A 32 that is slightly draggy on tarmac beats a 28 that is bouncing on farm track — gravel vs road goes into how to make that call.
• Old narrow rims: stop at 28. Fitting a 32 to a 15 mm internal rim gives you a tall, unsupported casing that squirms when you lean it over, and descending is not the place to discover that.

The shift from 23 to 28 was one of the few equipment changes in the last two decades that made almost everyone faster and more comfortable at the same time, with no real trade-off for a rider on matched rims. It happened slowly because it required admitting that a piece of received wisdom had been confounded from the start. That is usually how these things go.