Wondering how to run an FTP test without wrecking the result? You ride one steady, hard effort, take 95% of its average power, and train from that number. Pace it evenly and the figure is useful.
This page is part of the guide to training with power, and covers how to turn one test into usable numbers.
| Item | Value |
|---|---|
| Main test effort | 20 minutes, warm-up excluded |
| FTP calculation | Average power multiplied by 0.95 |
| Warm-up effort before the test | 5 minutes at maximum pace |
| Alternative multiplier some amateurs may suit | 92 or even 90 per cent |
| Upper limit of the threshold zone | Up to 105 per cent of FTP |
Figures come from the Favero test guide and the Cyclingnews feature. Protocols and advice can change, so check the sources before you rely on them.
What an FTP test actually measures
FTP stands for functional threshold power. In its simplest form it is the highest number of watts a rider can sustain aerobically, and it is widely measured over an hour, according to the Cyclingnews feature.
Riding at maximum power for an hour is incredibly difficult, so a shorter test stands in for it. Allen and Coggan suggested that the hour figure could be taken as 95 per cent of the average power in a 20-minute time trial.
That single number then sets your training intensities. The same article lists the zones as fractions of FTP:
- Active recovery: up to 55 per cent
- Endurance: up to 75 per cent
- Tempo: up to 90 per cent
- Threshold: up to 105 per cent
- VO2 max: up to 120 per cent
If a test result is wrong, every zone is wrong with it. That is why the setup and pacing matter more than the arithmetic.
How to run the 20-minute test step by step
The Favero guide describes the Allen-Coggan procedure as a warm-up followed by a best effort where you ride as hard as you can. Warm-up and cool-down are not counted in the test time.
- Ride 20 minutes easy.
- Do 3 x 1 minute fast-pedaling intervals up to a maximum of 120 rpm, with a minute at 80 rpm between them.
- Ride 5 minutes at a lower pace to recover.
- Ride 5 minutes at your maximum pace.
- Ride about five minutes at a low pace to recover, then start the 20-minute effort.
- Take the average power for the 20 minutes and multiply it by 0.95.
The guide says it is very important to include the all-out effort. Its stated reason is to reduce some of your freshness, so that 95% of the 20-minute number lands closer to what you might hold for the full hour.
Pacing and set-up that protect the result
Pacing is where most tests go wrong. The Cyclingnews feature warns that it is very easy to start too hard, which pushes you into VO2 max territory and leaves the second half of the test below threshold.
A few habits help:
- Pick your target from recent training and start slightly under it.
- Hold the effort even, and save anything left for the final minutes.
- Use the same bike, trainer or route, and power meter each time, so retests compare like with like.
- Test when rested, not at the end of a hard block.
Indoors on a trainer removes traffic and descents, which makes steady pacing easier. A flat road or steady climb can also work if it lets you hold a constant effort.
Is the 95% factor right for you?
The 95 per cent rule rests on one assumption: your hour power equals 95 per cent of your 20-minute power. The Cyclingnews feature notes that this is not always reliable, and that for many amateurs the drop-off between 20 and 60 minutes can be much greater.
Those riders might be better off using 92 or even 90 per cent, according to the same feature, because the 95 per cent rule sets their FTP too high. If your threshold sessions feel impossible, a too-generous multiplier is the first thing to suspect.
The right multiplier is a personal question. Compare how the zones feel over a few weeks of steady riding: if endurance work feels hard and threshold intervals are only just possible, the number is probably set too high.
Some riders also test on separate days and compare the averages. If the later effort is clearly higher, compare pacing, conditions and recovery before deciding which result is the better guide. Keep notes on how you felt towards the end of each attempt, because that tells you whether you started too hard or left too much unused.
Fuelling and sleep matter too. A test after a poor night, or on an empty stomach, will usually undersell your fitness, which then makes every later session too easy. Eat as you would before a hard race-day effort, and keep the day before light.
Research gives a sense of the scatter. A study in Frontiers in Physiology compared critical power with a 20-minute FTP and reported a mean bias of 7 ± 13 W. Treat any test result as an estimate with some give in it, not a fixed truth.
Other test formats
The Cyclingnews feature, published 28 September 2021, covers alternatives to the 20-minute effort. The ramp test, for instance, uses a series of regular upward steps in resistance ridden until failure, and the software then estimates FTP from the result.
Ramp tests are shorter and easier to pace, but the estimate comes from a formula that varies by platform. An FTP from one format is not directly interchangeable with one from another, so stick with one method when tracking progress.
Whichever format you choose, write down the date, the method and the result. A number without its method is hard to interpret later, and a change of format can look like a change in fitness when it is only a change in the formula.
Using your result in training
Once you have a number, set your zones from it. A fuller breakdown sits in the guide to cycling training zones, including how each band feels and what it trains.
The result also gives you a base for structured work such as sweet spot training, where intervals are prescribed as a share of FTP. If a session at the prescribed power is far too hard or too easy across several weeks, retest rather than guessing.
Retest after a block of consistent training, after a break, or whenever your sessions stop matching your zones. A power meter makes this easy, because the data you need is already being recorded.
Frequently asked questions
Do I have to do the 20-minute test?
Show answer
No. The Cyclingnews feature describes it as long-standing but hard to pace and based on an assumption that may not suit you, and it covers alternatives such as the ramp test.
Why do I do a 5-minute all-out effort first?
Show answer
The Favero guide says it reduces some of your freshness, so 95% of the 20-minute number is closer to what you might hold for the full hour.
What if the 95% result feels too high?
Show answer
Cyclingnews suggests some amateurs may be better off with 92 or even 90 per cent. Try a lower multiplier and see whether threshold sessions become manageable.
How should I calculate the result?
Show answer
Take the average power of the 20-minute effort and multiply it by 0.95. The product is an estimate of your FTP.
