What Is FTP, and How Should You Test It?

Short answer: FTP, or functional threshold power, is a cycling benchmark used to set your power-training zones. It is traditionally the highest power you can sustain for about an hour, but most riders estimate it from a 20-minute all-out effort and take 95% of the average. The result is a useful training estimate, not a precise measure of your physiology: how close it gets depends on how fit you are and how you run the test.

That matters because your zones, interval targets and the training load of each ride all hang off FTP. Training Stress Score (TSS), the usual measure of a session's load, is set so that one hour at FTP scores 100 (1). An FTP that is too high or too low skews both your workouts and your load figures.

What FTP is, and what it isn't

Hunter Allen and Andrew Coggan popularised FTP as a field-test number: the best average power you could hold for about an hour (1). The appeal was practical: no lab, no blood samples, just a power meter and one hard effort. Think of FTP as a reference mark on your training dashboard. It calibrates your zones; it doesn't pinpoint one boundary in your body. You don't need the lab terms below to use it, but research judges FTP against them, so they explain its limits.

FTP is meant to sit close to an important boundary in your physiology: the highest intensity at which your body can still settle into a steady state. Below that line, oxygen uptake and blood lactate eventually level off and you can keep going for a long time. Above it, they keep climbing until you have to stop.

Scientists call this boundary the maximal metabolic steady state (2, 3). Two lab measures are used to find it:

• Maximal lactate steady state (MLSS): the highest power at which blood lactate stays stable during a long, constant ride. Finding it takes several lab visits at different powers, with blood samples (4).
• Critical power (CP): the power that separates efforts your body can stabilise from efforts it can't. It is calculated from several all-out efforts of different lengths; above it, you can only keep going for a limited time (2).

How the two compare is still debated. One influential review argues that CP is the better marker of the true boundary (3). A more recent one finds the two can agree within a few watts when both are measured carefully (5). Lactate threshold is a related marker taken from a step test; labs define it in several ways, from the first sustained rise above baseline to the later point where lactate climbs steeply (6). FTP is defined by performance, what you can do for an hour, not by any of these lab measures.

How close is FTP to your threshold?

Often close, but not reliably for any one rider. A review of 15 studies on the 20-minute test found that those comparing FTP with lab markers reported gaps for individual riders too large to swap one for the other (7). The average gap also depends on the study and the threshold used.

The best case for FTP comes from a study of 15 trained and well-trained cyclists, where FTP and MLSS were almost the same on average (4). Even there, an individual rider's FTP could differ from their MLSS by roughly 9% in either direction (the range expected to cover 95% of riders). For a rider whose MLSS is 250 W, that is more than 20 W either way.

Other studies found FTP sitting above MLSS. In 11 trained cyclists who did their 20-minute test after a short warm-up, 95% of 20-minute power overestimated MLSS by about 12 W on average (8). In 18 well-trained men and women, MLSS was only about 93% of FTP (9).

In a study of 20 male cyclists, FTP was a good estimate of lactate threshold in the trained riders but significantly lower than it in the recreational ones (10).

Riding at FTP until exhaustion tells a similar story: 13 cyclists lasted about 34 minutes on average. Their oxygen uptake levelled off both at FTP and at 15 W above it. So FTP didn't mark the steady-state boundary, and the authors concluded it isn't a valid marker of the maximal metabolic steady state (11).

The same review (7) also found the 20-minute test reliable, meaning you get a similar number when you repeat it. Most of its studies were in trained male cyclists, so we know least about beginners and women.

The three common tests

The 60-minute test

The most direct test is an all-out hour. Few people do it: it is mentally brutal and hard to pace. In a study of 23 trained male cyclists, the 60-minute and 20-minute results were close on average but differed a lot for individual riders (12).

The 20-minute test

This is the standard field test. In Allen and Coggan's protocol, you ride a long warm-up of about 45 minutes, including easy riding, a few fast accelerations and a 5-minute hard effort. Then you ride 20 minutes as hard as you can and take 95% of the average power as your FTP (1, 13). If you average 250 W for the 20 minutes, your estimated FTP is about 238 W. That is an estimate, not a promise of what you could hold for an hour.

It is repeatable. In 25 trained cyclists tested a week apart, results varied by about 3% between tests (13).

The 95% factor is the weak point. A study of 120 cyclists who did both a 20-minute and a 60-minute test grouped riders by aerobic fitness (VO2max). The factor that best predicted hour power depended on the group (14):

• professionals: 96%
• well trained: 95%
• trained: 92%
• recreationally trained: 88%

A smaller study adds a caveat. In eight highly trained cyclists, 20-minute power tracked 60-minute power closely, yet hour power averaged about 90% of 20-minute power, not 95% (15). So 95% can overestimate hour power even in fit riders.

For less aerobically fit, recreational-level riders, then, 95% of 20-minute power probably overestimates hour power. That points the other way from the 20-cyclist study above, where recreational riders' FTP came out below their lactate threshold. The two used different yardsticks (a lab lactate marker versus actual hour power) and defined "recreational" differently, so neither settles it alone.

"About an hour" is also generous. In a study of road cyclists riding at their estimated FTP to exhaustion, recreational riders lasted a median of 35 minutes and professionals 51 minutes (16). Plan on FTP as a training anchor, not an hour you can actually hold.

Ramp tests and shorter tests

Many training apps use a ramp test instead. The power steps up every minute until you can't keep the pedals turning, and the app estimates FTP as a fixed share of your best minute; each platform sets its own share. It is quick. The catch is that on a rising ramp, oxygen uptake doesn't respond the way it does at a steady power (17). So one fixed percentage is unlikely to fit every rider equally well. Many coaches find that punchy riders, with a big anaerobic capacity and a strong sprint, get flattering ramp results and steady diesel types conservative ones, and that how fast the ramp rises matters too.

Shorter time trials of 4 or 8 minutes have also been tested, in small groups of trained riders (15, 18).

Why your number moves: the known biases

The warm-up. In a study of 21 trained cyclists, 20-minute power was about 5% lower after the long 45-minute warm-up with a 5-minute hard effort (around 256 W) than after shorter warm-ups (around 270 W) (19). Neither result is wrong. They are different protocols, so compare a test only with others done the same way.

Pacing. In that same study, riders started conservatively after the long warm-up and set off fast after the short ones (19), so pacing and warm-up can't be separated there. Most coaches advise an even effort.

Conditions. As a rule of thumb, heat drags your number down, so use a strong fan indoors. Don't compare indoor with outdoor results, or road-bike with time-trial-bike results; many riders, triathletes included, keep a separate number for each.

Sensitivity to change. In the 18-person study above, ten participants repeated the tests after seven months of training. Their MLSS rose by about 12 W on average, but their FTP didn't change (9). It is one small study, but a 20-minute test may miss modest gains in MLSS.

Noise. In a lab study of 11 trained cyclists that looked mainly at FTP versus MLSS, a change in 20-minute power smaller than about 9 W couldn't be told apart from normal test-to-test variation (8). At home, with your own power meter and conditions, the noise is likely larger. Repeat tests in trained riders varied by about 3% (13), so a change of a few per cent can be pure noise. Act on a change only if it is clearly larger than that and your other sessions agree.

Is critical power a better option?

For riders who want a number grounded in physiology, possibly, though CP has its own error: the result depends on the model and on the number and length of the efforts (20). That same review of power profiling recommends avoiding single-effort tests such as FTP and deriving CP from several efforts done under standard conditions (20).

In 17 trained cyclists and triathletes, CP was on average 7 W higher than FTP, and for 95% of riders the difference fell between about 19 W lower and 33 W higher (21). The authors advised not using them interchangeably.

The trade-off is effort: the study above used 12, 7 and 3 minutes, with 30 minutes' rest in between (21). Many training platforms, Tito included, still anchor power zones to FTP.

How often should you retest?

No study we know of has established the best interval, so this is coaching practice, not research. As a rule of thumb, retest every 6–8 weeks or at the end of a training block. After a long break, illness or injury, wait until you have had a week or two of easy and steady riding that feels normal, and never test while symptoms linger. For well-trained riders, testing more often rarely tells you much: lab test-to-test variation was about 9 W (8), close to the 12 W that seven months of training added to MLSS in one study of well-trained people (9). Newer riders, as many coaches find, often see changes sooner.

Between tests, your training gives you clues. Threshold intervals that feel noticeably easier at the same power, with a lower heart rate while you are fresh and sleeping well, suggest your FTP has risen. A lower heart rate with heavy legs or flat motivation is more often fatigue.

A simple FTP test protocol

What matters most: as a rule of thumb, especially for newer riders, a repeatable test and even pacing matter more than which protocol you pick. Treat your first result as provisional: the 95% factor may flatter you, while, in many coaches' experience, inexperience with pacing can drag the number down.

Before the test

1. Test fresh: an easy day or a rest day before, not at the end of a hard week.
2. Eat and drink as you would before a hard session, and don't test when unwell.
3. Use the same bike and power meter every time, calibrating it if you can. Indoors on a trainer is easiest to repeat; outdoors, pick a steady climb or road with no stops.

The test

1. Warm up for 20–45 minutes, mostly easy, with three or four short accelerations. Pick one warm-up and use it every time; if it includes the original protocol's 5-minute hard effort, keep that too.
2. Ride 20 minutes as hard as you can at an even pace. Without a number, start at an effort you could only just hold for 20 minutes: hard, about 8 out of 10, breathing deep and fast, able to say only a few words. If you have done this test before, your last 20-minute average is a pacing guide for the opening minutes, not a target to hold whatever happens; adjust by feel. Hold it steady, then push in the last 3–5 minutes. If you have to ease off sharply in the first five minutes, you started too hard.
3. Cool down for 10–15 minutes of easy spinning.
4. Take 95% of your 20-minute average as your starting estimate. It is a convention, not a rule that fits every rider. If you are less aerobically fit (often newer to structured training or riding only a few hours a week), consider 90% of your 20-minute average instead. That is a cautious training setting, not a validated formula: in the study of 120 cyclists, 92% best predicted hour power in the trained group and 88% in the recreational group (14).

Check it in training

Over the next week or two, ride a threshold workout. As a rule of thumb, two 15-minute blocks at or just under your new FTP with 5 minutes easy between should feel hard but controlled, about 7–8 out of 10.

What if the result seems wrong?

• You can't finish both blocks, or the workout feels close to all-out, even though you went in fresh: as a rule of thumb, lower FTP by 3–5% and check again.
• The check felt well below 7 out of 10 and threshold sessions stay easy for two weeks: your first test may have been held back by pacing; retest.

No power meter? Friel's heart-rate test is a widely used coaching alternative. Warm up as for the power test, then ride 30 minutes alone as hard as you can; the average heart rate of the last 20 minutes is your lactate threshold heart rate (LTHR) (22). Use it to set heart-rate zones, but as a rougher guide than FTP: heart rate shifts with heat, hydration and fatigue.

Who should be careful: a maximal test puts real strain on your heart. If you have a heart condition, are returning from illness, or are new to hard exercise, check with your doctor first. Stop straight away if you feel chest pain, unusual breathlessness, dizziness or faintness.

References

1. Allen H, Coggan AR, McGregor S. Training and Racing with a Power Meter. 3rd ed. VeloPress; 2019.
2. Poole DC, Burnley M, Vanhatalo A, et al. Critical Power: An Important Fatigue Threshold in Exercise Physiology. Med Sci Sports Exerc. 2016. doi:10.1249/mss.0000000000000939
3. Jones AM, Burnley M, Black MI, et al. The maximal metabolic steady state: redefining the 'gold standard'. Physiol Rep. 2019. doi:10.14814/phy2.14098
4. Klitzke Borszcz F, Ferreira Tramontin A, Pereira Costa V. Is the Functional Threshold Power Interchangeable With the Maximal Lactate Steady State in Trained Cyclists? Int J Sports Physiol Perform. 2019. doi:10.1123/ijspp.2018-0572
5. Caen K, Poole DC, Vanhatalo A, et al. Critical Power and Maximal Lactate Steady State in Cycling: "Watts" the Difference? Sports Med. 2024. doi:10.1007/s40279-024-02075-4
6. Jamnick NA, Pettitt RW, Granata C, et al. An Examination and Critique of Current Methods to Determine Exercise Intensity. Sports Med. 2020. doi:10.1007/s40279-020-01322-8
7. Mackey J, Horner K. What is known about the FTP20 test related to cycling? A scoping review. J Sports Sci. 2021. doi:10.1080/02640414.2021.1955515
8. Lillo-Beviá JR, Courel-Ibáñez J, Cerezuela-Espejo V, et al. Is the Functional Threshold Power a Valid Metric to Estimate the Maximal Lactate Steady State in Cyclists? J Strength Cond Res. 2022. doi:10.1519/jsc.0000000000003403
9. Inglis EC, Iannetta D, Passfield L, et al. Maximal Lactate Steady State Versus the 20-Minute Functional Threshold Power Test in Well-Trained Individuals: "Watts" the Big Deal? Int J Sports Physiol Perform. 2020. doi:10.1123/ijspp.2019-0214
10. Valenzuela PL, Morales JS, Foster C, et al. Is the Functional Threshold Power a Valid Surrogate of the Lactate Threshold? Int J Sports Physiol Perform. 2018. doi:10.1123/ijspp.2018-0008
11. Wong S, Burnley M, Mauger A, et al. Functional threshold power is not a valid marker of the maximal metabolic steady state. J Sports Sci. 2022. doi:10.1080/02640414.2023.2176045
12. Borszcz FK, Tramontin AF, Bossi AH, et al. Functional Threshold Power in Cyclists: Validity of the Concept and Physiological Responses. Int J Sports Med. 2018. doi:10.1055/s-0044-101546
13. Klitzke Borszcz F, Tramontin AF, Costa VP. Reliability of the Functional Threshold Power in Competitive Cyclists. Int J Sports Med. 2020. doi:10.1055/a-1018-1965
14. Sitko S, Cirer-Sastre R, López-Laval I. An Update Of The Allen & Coggan Equation To Predict 60-Min Power Output In Cyclists Of Different Performance Levels. Int J Sports Med. 2023. doi:10.1055/a-2079-1363
15. MacInnis MJ, Thomas ACQ, Phillips SM. The Reliability of 4-Minute and 20-Minute Time Trials and Their Relationships to Functional Threshold Power in Trained Cyclists. Int J Sports Physiol Perform. 2019. doi:10.1123/ijspp.2018-0100
16. Sitko S, Cirer-Sastre R, López-Laval I. Time to exhaustion at estimated functional threshold power in road cyclists of different performance levels. J Sci Med Sport. 2022. doi:10.1016/j.jsams.2022.06.007
17. Keir DA, Paterson DH, Kowalchuk JM, et al. Using ramp-incremental V̇O2 responses for constant-intensity exercise selection. Appl Physiol Nutr Metab. 2018. doi:10.1139/apnm-2017-0826
18. Gavin TP, Van Meter JB, Brophy PM, et al. Comparison of a field-based test to estimate functional threshold power and power output at lactate threshold. J Strength Cond Res. 2012. doi:10.1519/jsc.0b013e318220b4eb
19. Tramontin AF, Borszcz FK, Costa V. Functional Threshold Power Estimated from a 20-minute Time-trial Test is Warm-up-dependent. Int J Sports Med. 2022. doi:10.1055/a-1524-2312
20. Leo P, Spragg J, Podlogar T, et al. Power profiling and the power-duration relationship in cycling: a narrative review. Eur J Appl Physiol. 2022. doi:10.1007/s00421-021-04833-y
21. Karsten B, Petrigna L, Klose A, et al. Relationship Between the Critical Power Test and a 20-min Functional Threshold Power Test in Cycling. Front Physiol. 2021. doi:10.3389/fphys.2020.613151
22. Friel J. The Triathlete's Training Bible. 4th ed. VeloPress; 2016.