Repeated sprint ability: why the fifth effort is the one that matters

Ask a rider how good their sprint is and they will tell you a number: 1,250 W, 1,400 W, whatever their head unit recorded on a downhill with a tailwind. It is the wrong number. In an actual race that peak is produced once, at the end, after two hours of riding, and it is usually well below what the same rider can do fresh in a car park.

What decides races is what happens on the fourth and fifth effort. The bunch splits over a rise, comes back together, splits again in the next corner, and someone attacks off the front of it. The rider who wins is rarely the one with the highest single sprint. It is the one whose fifth sprint is still nearly as good as their first.

That quality has a name — repeated sprint ability — a fairly well understood physiological basis, and a training implication that surprises most sprinters: the thing that most limits your fifth effort is aerobic.

What a race actually asks

Power analysis of a world-class professional sprinter across grand tours showed something that should reframe how amateurs think about this. The sprint itself lasted roughly 14 seconds at around 1,000 W — impressive, but noticeably below what a sprinter of that calibre produces rested. More telling was what came before it: the final ten minutes into the finish were ridden at very high power, and the sprint was launched from the end of that.

The same group looked at the other way races are decided — getting away from the bunch — and found the same structure at a different timescale. Establishing a breakaway is not one big effort. It is a series of very hard efforts with almost no recovery between them, and the power required to make it stick is concentrated in the first minutes rather than spread evenly. A sprint and a breakaway are the same problem wearing different clothes: repeated maximal work on an incomplete recovery clock.

The same logic scales all the way down. In a fourth-category criterium nobody is producing 1,500 W, but the accordion effect out of every corner produces the same structure: short maximal efforts separated by recovery that is never quite enough. Your first criterium describes the experience from the inside; this post is about the physiology underneath it.

The number that matters is the decrement

Repeated-sprint tests all share a structure: several maximal efforts of a few seconds each, separated by recovery too short to complete. What comes out is not one number but three — your best effort, your mean effort, and the percentage by which you fell off.

One caution before you go and test yourself. The best-sprint and mean-sprint figures are reasonably repeatable. The decrement score — the percentage drop — is noisier, because it is a ratio of two measurements each with their own error. Treat a change of one or two percentage points between tests as nothing. Treat a change of five as possibly real, and only if the warm-up, the gearing and the time of day were the same.

Why the fifth effort fades

Within a single six-second maximal sprint, most of the energy comes from stored phosphocreatine and anaerobic glycolysis. Phosphocreatine is the fast fuel — it is depleted substantially within seconds, and everything about your ability to repeat the effort depends on how quickly you put it back.

That restoration is an aerobic process. It requires oxygen delivered to the muscle, and its rate is directly dependent on oxygen availability: studies manipulating inspired oxygen have shown phosphocreatine recovery slowing in hypoxia and speeding up in hyperoxia. Recovery of power output after a 30-second sprint tracks the recovery of muscle phosphocreatine closely.

And as the sprints accumulate, the fuel mix shifts. In a series of four 30-second maximal sprints, the aerobic contribution to each bout rose substantially from the first to the last, while total power output fell. By the fourth effort a meaningful share of the work is being done aerobically — in a sprint. That is the heart of the argument.

How strong is the aerobic link, honestly

This is where the literature deserves a careful reading rather than a slogan. The mechanism is not in doubt: phosphocreatine resynthesis is oxidative, and it is limited by oxygen availability. What is less clean is how tightly that translates into a correlation between someone's VO2 max and their sprint decrement.

Reviews of this question report a relationship that is real but moderate, and inconsistent across studies. Some find clear associations between aerobic fitness and recovery from intermittent high-intensity work; others find little. The comprehensive reviews of repeated-sprint ability treat aerobic fitness as one contributing factor among several — alongside muscle buffering capacity, oxygen uptake kinetics, and neural drive — rather than as the single determinant.

So the honest version is this: a bigger aerobic engine helps you repeat sprints, the mechanism is clear, and the size of the effect in any individual is not predictable from their VO2 max alone. It is still the most reliable lever an amateur has, because it is also the one that improves everything else. The case for base miles is not only about long rides.

Training it

The published recommendations for training repeated-sprint ability say to work on both ends: the ability to produce a single high-quality effort, and the ability to recover between them. Those are different sessions and they do not substitute for each other. A block of pure sprint work raises the peak and often makes the decrement worse, because you have raised the number everything is being compared against without touching the recovery machinery.

If you want to see this in your own data rather than take it on trust, the useful view is your short-duration power across a season — Moveee's power profile fits your best efforts at each duration from your own rides, and a sprinter who has been doing the aerobic work properly will see the five-second number hold while the one-minute and five-minute numbers climb underneath it. That is the shape you want.

How this differs from the sessions you already do

• Not the same as sprint training. Classic sprint work uses full recovery precisely so that every effort is maximal. That builds peak power and technique — how to build a sprint that actually wins things covers it, and neuromuscular sprint training has the sessions themselves, including a repeatability set. Repeated-sprint work deliberately denies you that recovery, and the sessions feel completely different.
• Not the same as 30/30s. A 30/30 session is a VO2 max session: the efforts are hard but submaximal, and the target is time spent near maximal oxygen uptake. Repeated sprints are all-out from the first pedal stroke, much shorter, and target a different limiter entirely.
• Not the same as W′ work. Your anaerobic work capacity sets how much you can spend above threshold in total. Repeated-sprint ability is about how fast you can refill it between withdrawals. W′ and anaerobic capacity is the companion piece.

What to take away

If you race anything with corners, hills or attacks, stop measuring your sprint by its peak. Do a repeated-sprint set twice in a build block, six weeks apart, on the same drag, and compare the last effort rather than the first. That single number tells you more about whether you will be in the finish than your maximum watts ever will.

And then do the unglamorous thing. The sprinter who can sprint twice is not the one who did more sprints. It is the one who spent the winter building the aerobic system that puts the phosphocreatine back.