Cycling economy: why the same watts cost two riders different amounts

Two riders sit on adjacent turbo trainers, both holding 250 W. Both are breathing hard. One is consuming meaningfully more oxygen than the other to produce exactly the same number on the screen. Nothing is wrong with either of them — they simply differ in how much of the chemical energy they burn ends up turning the cranks.

That ratio is called gross mechanical efficiency, and in cyclists it lands somewhere around 18 to 24 percent. The rest leaves as heat. It sounds like an obscure laboratory quantity, and for the purposes of your Tuesday session it largely is — but it is one of the few variables that can separate two riders with identical VO2max and identical threshold power, and the longer the event, the more it matters.

Where the energy actually goes

Start with the arithmetic, because it makes the rest obvious. If you are producing 250 W at the pedals at a gross efficiency of 20 percent, your body is burning energy at five times that rate.

That second panel is the whole reason anyone cares. Going from 20 to 21 percent gross efficiency is a five percent increase in power at the same oxygen cost. For a rider at 250 W that is twelve watts, which is more than most people get from a training block. Flip the arithmetic round and it looks equally good: hold 250 W, and one extra point of efficiency drops your oxygen demand from about 3.6 to about 3.4 litres a minute.

Efficiency is the quiet multiplier sitting behind every other number. Your VO2max sets how much oxygen you can deliver; efficiency sets how many watts you get for it. That is why it is possible for a rider with an unremarkable VO2 max to hold power that the number says they should not.

Gross, net, delta — and why the distinction matters

Three definitions get used more or less interchangeably in conversation and mean quite different things in a laboratory.

• Gross efficiency is work done divided by total energy expended. Simple, and the one most often reported. It includes the cost of simply being alive and sitting upright, so it rises as the work rate rises, which makes comparisons across intensities awkward.
• Net efficiency subtracts resting metabolic rate first. Slightly more physiologically honest, slightly harder to measure well.
• Delta efficiency is the slope of energy expenditure against work rate across several steps — in principle the closest thing to "the efficiency of the working muscle". In practice it is calculated from a handful of data points, and small errors in each of them produce large errors in the slope.

The reason to know this is that claims about efficiency improving are usually claims about one of these three specifically, and they do not always move together. A review of the field concluded that much of the apparent disagreement in the literature comes down to which definition was used and at what work rate it was measured.

What actually determines it

The fibre-type finding is worth dwelling on because it connects two things that otherwise look unrelated. Slow-twitch fibres are more economical at the contraction speeds cycling uses, so a rider with more of them gets more watts per litre of oxygen. That same rider probably also has a higher fractional utilisation and better fatigue resistance, which is why efficiency and "diesel engine" tend to travel together. It also means a large part of your efficiency was decided long before you bought a bike.

Cadence is the one lever you can pull today, and the finding is counterintuitive: efficiency is generally highest at cadences lower than riders choose. That does not mean you should grind everywhere. Freely chosen cadence appears to optimise something other than pure metabolic cost — perceived effort, muscle force, fatigue resistance over hours — and our piece on cadence covers why the efficient cadence and the sensible cadence are not the same number.

Can you train it?

Probably a little. Certainly not as much as the enthusiastic version of this story suggests.

The strongest supporting evidence is longitudinal: gross efficiency measured in trained cyclists rose across a competitive season, and world-class professionals followed over several years showed improvement. There is also a clear cross-sectional gap between trained and untrained people that survives adjustment for age.

The strongest counter-evidence is the measurement problem. When the reliability of cycling efficiency was tested directly, day-to-day variation turned out to be of a similar order to the changes being reported as training effects. A season-long improvement of half a percentage point is real if it is real, and indistinguishable from drift if it is not.

The Tour de France argument

The most famous single claim in this literature is worth telling honestly, because it shows how thin the ground is. In 2005 a paper reported that a Tour de France champion's muscular efficiency had improved by around eight percent over seven years of testing, framed as an explanation for his transformation as a rider. In 2008 another group published a direct rebuttal in the same journal, arguing that the delta efficiency calculation in that paper was simply wrong, and a reply followed.

The dispute was never cleanly settled. What it demonstrates is that when four people who study this for a living cannot agree on the efficiency of one extensively tested athlete, you should not be confident about half a point of change in yourself measured once.

Why it matters more the longer you ride

Over a 20-minute climb, efficiency shows up as watts and nothing else. Over six hours it shows up three separate times.

First, as fuel. A more efficient rider burns less total energy for the same work, which means fewer grams of carbohydrate to get in per hour and a smaller gap between what you can absorb and what you are spending. Second, as heat. Eighty percent of your metabolic power leaves as heat, so a less efficient rider is generating more of it at the same speed and paying for it in cardiovascular strain on a hot day. Third, as durability — the slow decay of your power over hours, which we cover in why your power at hour four is the number that matters.

None of this is something you can measure without a metabolic cart and a mask, and no consumer device estimates it honestly. What you can do is act on the determinants that are actually under your control: accumulate the aerobic training history that appears to shift it, which is the unglamorous case made in the evidence for volume, and stop assuming that a higher cadence is automatically a better one.

The short version

Efficiency explains why watts are not a complete description of an athlete. It is substantially determined by muscle you were largely born with, it moves slowly and modestly with training, and it is measured with enough error that individual claims about it deserve scepticism. A single percentage point is worth about five percent of your power, which is why it is interesting — and why the temptation to overclaim it has been so hard for the field to resist.