Tri ToolkitCalculators
Bike › Guide

How to Do an FTP Test: 20-Minute and Ramp

Updated 2026-11-21 · draft, awaiting author review

The simplest FTP test is a 20-minute time trial: warm up well, ride 20 minutes as hard as you can hold evenly, and multiply your average power by 0.95. A ramp test is shorter: the target rises every minute until you stop, and FTP is 75% of your best minute. Either works if you use the same test, in the same conditions, every time.

What an FTP test tells you

Functional Threshold Power (FTP) is roughly the highest power you can hold for about an hour. You use it to set your training zones and your race watts. Few people want to ride flat out for an hour, so most tests are shortcuts: a shorter effort and a factor that scales it to an hour.

That is also the catch. The factor is an average for a group. At the group level, FTP from the 20-minute test sits close to laboratory thresholds. For you as an individual, it can be several percent off in either direction Moderate evidence. Treat your FTP as a useful training number, not a lab measurement.

Two tests at a glance

Both tests end with you riding as hard as you can. The difference is how you get there: the 20-minute test asks you to pace a long, steady effort, while the ramp test makes the pacing decisions for you and simply gets harder until you cannot continue.

Power over time in a 20-minute FTP test and a ramp testTwo schematic power profiles in percent of FTP. Top: the 20-minute test. About 45 minutes of warm-up, building from easy riding, with one 5-minute hard effort and then easy riding, followed by 20 minutes held steady at about 105 percent of FTP, because FTP is 95 percent of 20-minute power. Then a cool-down. Bottom: the ramp test. 5 minutes easy, then the target rises every minute until you cannot hold it, usually within about 20 minutes. Your best minute is about 133 percent of FTP, because FTP is 75 percent of best 1-minute power. Then a cool-down.20-minute test (Allen and Coggan): about 80 min in total50%100%warm-up5 min hardeasy20 min steady: about 105% of FTPcool-down0 min20 min40 min60 min80 minRamp test (TrainerRoad, Zwift): about 25 min in total50%100%easyone step up every minutebest minute: about 133% of FTP, then stopcool-down0 min5 min10 min15 min20 min25 minVertical axis: percent of FTP. Schematic: your own warm-up, steps and stopping point will differ.
Schematic. The 105% and 133% follow from the formulas (FTP = 0.95 × 20-minute power; FTP = 0.75 × best 1-minute power in a ramp). Warm-up after the standard protocol as described by Borszcz et al. 2020; ramp steps of 20 watts a minute as Zwift describes them; the length of the easy start is schematic. Both tests end in a hard finish, but only the 20-minute test asks you to pace.

The 20-minute test, step by step

The 20-minute test comes from Hunter Allen and Andrew Coggan’s book Training and Racing with a Power Meter: FTP is estimated as 95% of your average power over a 20-minute all-out time trial. The standard version starts with a 45-minute warm-up with easy riding, fast accelerations and a 5-minute time trial Moderate evidence.

A practical version Coaching practice:

  1. Warm up for about 20 minutes, building from easy riding to a moderate effort.
  2. Ride 5 minutes hard, roughly at the best pace you could hold for those five minutes.
  3. Ride easy for about 10–15 minutes until your breathing is back to normal.
  4. Ride 20 minutes as hard as you can hold evenly. Start a little below what you think you can manage, settle in for the middle part, and use what is left in the last 3–5 minutes.
  5. Cool down for 10 minutes or more.
  6. Multiply the 20-minute average by 0.95. A 20-minute average of 240 watts gives an FTP of 228 watts.

How much the warm-up matters for the result is less clear than it sounds. In a study of 15 cyclists who did the 20-minute test after the standard warm-up, a short 10-minute warm-up or no warm-up, FTP was closely related to a laboratory threshold in all three cases Limited evidence. What matters more is that you do the same warm-up every time, so changes reflect fitness rather than routine. The 20-minute effort itself is repeatable: in a reliability study of competitive cyclists the test varied by about 3% from one attempt to the next, while the warm-up varied more Moderate evidence.

Enter your 20-minute average in the zones calculator, choose “20-minute test”, and it does the multiplication and gives you power zones and watts per kilo. Saved once, the number fills in across the site’s other calculators.

The ramp test

The ramp test is TrainerRoad’s default test, and Zwift uses the same rule. It is a manufacturer method, not a protocol from research Coaching practice:

  1. Start with a short, easy warm-up.
  2. The target power rises in steps. On Zwift’s standard ramp it goes up by 20 watts every minute. On a smart trainer it is set for you.
  3. Keep going until you can no longer hold the target. You do not need to finish the step you are on.
  4. FTP is 75% of your best 1-minute power during the test.

TrainerRoad says it uses your best minute rather than your last, so easing off after you fail does not change the result. If you were above target in the final minute, the app trims the result slightly.

The ramp test is short, needs almost no pacing skill and is less draining than 20 minutes at your limit. Its weakness is the fixed 75%. Your best minute in a ramp depends a lot on your anaerobic capacity, and Zwift Insider notes that riders with a high anaerobic power relative to their aerobic power may get an FTP that is too high Coaching practice. TrainerRoad, for its part, recommends the ramp test for triathletes and time triallists too, arguing that a poorly paced 20-minute test is less accurate. If you have used the zones calculator before, pick “Ramp test” and enter your best minute.

Other protocols: 8, 60 and two efforts

8-minute test. Developed by Chris Carmichael’s coaching company (CTS) with Edmund Burke: two all-out 8-minute time trials with 10 minutes of easy spinning between them. FTP is 90% of the higher of the two averages Coaching practice. CTS says it suits riders who are not yet good at pacing, and notes that an 8-minute effort has a substantial anaerobic component, so like the ramp it can flatter riders with a big anaerobic capacity.

60-minute test. The definition itself: your best average power for an hour. It is the most direct and the most demanding, and it is hard to pace and recover from. Few age-groupers do it, and most protocols exist precisely to avoid it.

Why not just trust the 20-minute number for an hour? Because many riders cannot hold it. When 23 trained cyclists rode to exhaustion at their FTP from the 20-minute test, they lasted on average about 51 minutes, with a wide spread of plus or minus about 16 minutes Moderate evidence. The 20-minute FTP, the 60-minute FTP and a laboratory threshold agreed on average, but the limits of agreement between them reached about 60 watts in either direction.

How accurate is the 95% rule?

The 95% factor works reasonably well on average and less well for individuals. The studies, all on fairly small groups of mostly male, trained cyclists:

Study Riders What they found
Borszcz et al. 2019 15 trained and well-trained FTP was on average 1.4% from the maximal lactate steady state, but with individual errors of about ±9%
Inglis et al. 2020 18 well-trained The maximal lactate steady state was on average 93% of the 95% FTP, so the formula overestimated it; changes with training were not mirrored by FTP
Lillo-Beviá et al. 2022 11 cyclists and triathletes The 95% factor overestimated by about 12 watts on average; 91% fitted this group better
Valenzuela et al. 2018 20 recreational and trained FTP was a good estimate of lactate threshold in trained cyclists, less so in recreational riders; the error depended on fitness
Borszcz et al. 2018 23 trained 20-minute and 60-minute FTP agreed on average, but should not be used interchangeably for individuals
Wong et al. 2022 13 cyclists Riders lasted about 34 minutes at FTP; oxygen uptake suggests FTP is not a true boundary between heavy and severe exercise

What this means in practice Moderate evidence:

  • Your FTP is an estimate with a margin of a few percent, and it is more likely to be a little high than low.
  • No single factor fits everyone. Some riders hold 95% of their 20-minute power for close to an hour; others fade well before.
  • The test is still useful for tracking change, as long as you repeat the same protocol.

If your threshold sessions feel impossible, or you blow up at your planned race power, assume your FTP is set too high and drop it a few percent rather than forcing the workouts. Our guide to training zones explained covers how much the different threshold systems disagree Coaching practice.

Indoors or outdoors?

Average power in a 20-minute test was almost the same indoors and outdoors in a study of 20 competitive cyclists: 280 watts on an indoor trainer and 284 watts outdoors, with the two strongly related Moderate evidence. Outdoors, power varied about twice as much from second to second, because of hills, corners, wind and traffic.

That makes indoors the easier place to test well: no stops, no traffic, and the same set-up every time. Outdoors works if you have a safe, flat or steady uphill stretch long enough for 20 minutes without junctions. Many riders find their outdoor power a little different from their indoor power, so test where you do most of your structured training, and do not compare indoor and outdoor results directly Coaching practice.

Indoors, heat is the big variable. Without airflow your body has no wind to cool it, heart rate drifts up and power fades. Use a strong fan, keep the room cool and have a drink at hand. Research on fans at room temperature is limited, but in hot conditions airflow clearly matters, as our guide to racing in the heat explains Limited evidence.

How often to test

TrainerRoad recommends testing every four to six weeks during a structured plan Coaching practice. If you do not follow a plan, every 6–10 weeks is enough, or sooner after a clear change: an illness, a long break or a big jump in fitness.

Testing more often than that rarely helps. A hard test costs a training day, you need to be fresh for it, and day-to-day noise of a few percent can hide real progress. Two signs it is time to retest: threshold intervals have started to feel clearly easier, or you can no longer finish them.

Mistakes that skew the number

Starting too hard. The most common way to ruin a 20-minute test. If the first five minutes are well above your final average, you will fade. Hold back early and finish strong Coaching practice.

Testing tired. A test after a hard week measures fatigue, not fitness. Have an easy day or two before it, eat normally, and test at the same time of day each time.

A hot room. See above: a fan and a cool room are part of the protocol.

Power meter not zeroed. Do a zero offset (often called calibration) before the test, the way your power meter’s maker describes it, and run your smart trainer’s own calibration routine. Use the same device for every test; two power meters rarely read exactly the same Coaching practice.

Changing the protocol. A ramp result and a 20-minute result are not the same measurement. Switching between them, or changing the warm-up, makes changes in FTP hard to read.

ERG mode on a 20-minute test. With the trainer holding a fixed power you cannot pace by feel, and a target set too high ends the test early. Use resistance or slope mode for self-paced tests.

Safety note: an FTP test is a maximal effort. Only test when you are healthy, and stop if you feel chest pain, dizziness or unusual shortness of breath. This is general information for healthy adults, not medical advice.

Using your FTP

Your FTP sets your training zones and your race power. Most age-groupers ride a 70.3 bike at about 80–85% of FTP and an Ironman bike at about 65–75%, so a 228-watt FTP points to roughly 182–194 watts for a 70.3 and 148–171 watts for an Ironman Coaching practice. The race pacing planner works out your own range from FTP and expected bike time, and our guide to pacing a 70.3 or an Ironman explains why the numbers sit that low.

Because a too-high FTP makes every race target too high, the margin of error matters more on race day than in training. If your test is old or you suspect it flattered you, use the lower end.

If you train more

Riders on 12 or more hours a week, aiming for a podium or a slot, can get more out of testing:

  • Test as part of the plan, not as an event. Put a test at the start of each block, and use key sessions in between as checks. A long threshold set or a race at known effort tells you a lot about whether your FTP is right Coaching practice.
  • Try a critical power test. Critical power (CP) comes from several all-out efforts of different lengths. In one study of 17 trained cyclists and triathletes, CP from 12-, 7- and 3-minute time trials with 30 minutes of rest between them averaged 256 watts against 249 watts for the 20-minute FTP. The two were closely related, but individual differences of up to about 33 watts meant they should not be used interchangeably Moderate evidence. The authors point to studies showing that CP marks the boundary between steady and non-steady metabolism, and the efforts also give you a picture of your short-duration power.
  • Compare FTP with your race files. If you regularly hold your planned Ironman power and run well off the bike, your FTP and pacing fit. If not, the problem may be the number, not your legs.
  • Watts per kilo vs watts. FTP divided by body mass matters most on climbs. On flat and rolling courses, absolute watts and your aerodynamics decide your speed, which is why a heavier rider with a good position can be fast on a flat course. Our guide to aero upgrades shows how much time position is worth.

For the structure around the tests, see our guide to training 12–16 hours a week for a 70.3 or Ironman.

FAQ

Ramp test or 20-minute test?

Both are reasonable. The ramp is shorter and needs no pacing skill, while the 20-minute test is closer to the steady efforts of a triathlon and is less affected by sprint power. Pick one and stick with it, so your results are comparable.

Is 95% of 20-minute power accurate?

On average it is close, but for individuals it can be several percent off. Several studies found it tends to overestimate the highest steady-state power, so if your FTP feels too high in training, lower it a few percent.

How often should I do an FTP test?

Every four to six weeks during a structured plan, or every 6–10 weeks otherwise. Retest sooner after illness, a long break or a clear change in fitness.

Is FTP lower indoors?

In one study of 20 competitive cyclists, average power in a 20-minute test was almost the same indoors and outdoors. Heat is the main reason people test lower indoors, so use a strong fan and a cool room.

Why is my ramp FTP higher than my 20-minute FTP?

The ramp result depends on your best minute, which is largely anaerobic. Riders with a strong sprint or a big anaerobic capacity often get a higher ramp FTP than they can hold for long efforts.

Can I do an FTP test without a power meter?

You need power data from a power meter or a smart trainer to get a number in watts. Without one, use a heart-rate test or effort by feel to set your zones, and treat speed outdoors as too dependent on wind and terrain to use.

Should I test in ERG mode?

Not for the 20-minute or 8-minute test, because you need to pace by feel. A ramp test on a smart trainer sets the target power for you by design.

Sources

  1. Borszcz et al. 2018, Functional threshold power in cyclists: validity of the concept and physiological responses, Int J Sports Med
  2. Borszcz et al. 2020, Reliability of the functional threshold power in competitive cyclists, Int J Sports Med
  3. Valenzuela et al. 2018, Is the functional threshold power a valid surrogate of the lactate threshold?, IJSPP
  4. Borszcz et al. 2019, Is FTP interchangeable with the maximal lactate steady state in trained cyclists?, IJSPP
  5. Inglis et al. 2020, Maximal lactate steady state versus the 20-minute FTP test, IJSPP
  6. Lillo-Beviá et al. 2022, Is FTP a valid metric to estimate the maximal lactate steady state?, J Strength Cond Res
  7. Barranco-Gil et al. 2020, FTP, respiratory compensation point and warm-up protocols, IJSPP
  8. Karsten et al. 2021, Critical power test vs 20-min FTP test in cycling, Front Physiol
  9. Wong et al. 2022, Functional threshold power is not a valid marker of the maximal metabolic steady state, J Sports Sci
  10. Jeffries et al. 2019, Variability in power output during indoor and outdoor time trials, IJSPP
  11. TrainerRoad, Ramp Test FAQs (manufacturer method)
  12. TrainerRoad, How to test with the Ramp Test (manufacturer)
  13. TrainerRoad blog, How often should I take an FTP test? (manufacturer)
  14. CTS / TrainRight, Why two 8-minute efforts instead of one 20-minute effort (test author)
  15. CTS / TrainRight, CTS Field Test: how to calculate training zones (test author)
  16. Zwift, How to measure your cycling fitness effectively (manufacturer)
  17. Zwift Insider, How (and why) to take an FTP test on Zwift