FTP — functional threshold power — is the single most-used number in cycling training. It sets your zones, anchors your intervals, determines your pacing on race day, and provides the baseline against which every future test is measured. But the number is only as good as the test that produced it, and there are four credible ways to get there, each with its own conversion factor and its own failure modes.
This is a side-by-side comparison of the 20-minute test, the ramp test, the 2×8-minute test, and the 60-minute test — what each one measures, what it misses, and which one suits your physiology. If you want to run the numbers after testing, the FTP Test Calculator does the maths for any protocol.
Contents
- What FTP Actually Is
- The 20-Minute Test
- The Ramp Test
- The 2×8-Minute Test
- The 60-Minute Test
- Which Test Suits Which Rider
- Common Testing Mistakes
- How Often to Test
- Consistency Beats Protocol Choice
What FTP Actually Is
The concept comes from Dr. Andrew Coggan and Hunter Allen, who formalised it in Training and Racing with a Power Meter. FTP is defined as the highest power a rider can sustain in a quasi-steady state for approximately one hour. "Quasi-steady state" is the key phrase — it means the effort is hard enough that lactate is accumulating, but not so hard that accumulation outpaces clearance. You are riding right at the boundary. Step above it and lactate starts to stack. Step below it and you could hold the effort significantly longer.
Physiologically, FTP approximates what exercise scientists call maximal lactate steady state (MLSS) — the highest intensity at which blood lactate concentration remains stable over time. It is not a perfect proxy. Lab-measured MLSS and field-measured FTP do not always agree, and the relationship shifts with fibre composition, fatigue state, and test duration. But for setting training zones and tracking fitness over time, FTP is close enough. It remains the default reference point for structured cycling training worldwide.
The problem is that nobody wants to ride flat out for sixty minutes just to set their zones. It is brutal, it requires perfect pacing, and most riders produce a lower number from a poorly paced hour-long effort than from a well-paced shorter test with a conversion factor. That is why shorter protocols exist — each estimates your one-hour capacity from a more manageable effort using a different multiplication factor.
The 20-Minute Test
Protocol: Warm up thoroughly (15-20 minutes including several progressive efforts), then ride as hard as you can sustain for exactly 20 minutes. Record average power. Multiply by 0.95.
Why 0.95? The factor accounts for the anaerobic contribution that inflates a 20-minute effort beyond what you could hold for a full hour. Over 20 minutes, your anaerobic system contributes roughly 5% more power than it would at a truly sustainable pace. Coggan and Allen calibrated the factor against actual 60-minute efforts in trained cyclists, and it has since been validated across a wide range of abilities.
When it works: The 20-minute test is the most widely used field protocol for good reason. It captures sustained aerobic capacity — you cannot bluff your way through 20 minutes on anaerobic reserves alone. By the five-minute mark, your aerobic system is doing the vast majority of the work. The protocol has been standardised long enough that comparing results against population data is meaningful, and for experienced riders who know how to pace it, the 20-minute test is the most reliable option available outside a lab.
When it lies: Pacing is everything. The classic mistake is going out at a power you cannot sustain, holding it for six or seven minutes on adrenaline and stubbornness, and then fading badly over the second half. A rider who averages 280 watts for 20 minutes with even pacing and a rider who averages 280 watts via a 310-watt first half and a 250-watt second half have not done the same test. The evenly paced effort is a legitimate threshold approximation. The poorly paced one is a maximal effort with a long anaerobic tail, and the 95% factor does not correct for that.
The test also lies for riders whose anaerobic contribution over 20 minutes deviates significantly from the assumed 5%. A rider with an unusually strong anaerobic engine — someone who can time-trial at 105% of their true threshold for 20 minutes before blowing up at 25 — will produce an inflated number. The 95% factor assumes a population-average anaerobic contribution. You may not be average.
The warm-up matters more than you think. A proper warm-up for a 20-minute test includes at least 15 minutes of progressive riding, two or three 1-minute efforts at increasing intensity, and a brief rest before the test begins. The goal is to open up your aerobic system and pre-activate your neuromuscular pathways without creating fatigue. An inadequate warm-up suppresses your result. An overly aggressive warm-up fatigues you before you start. Most guides understate how much the warm-up affects the outcome.
The Ramp Test
Protocol: Start at a low wattage (often 100W) and increase by a fixed amount — typically 20 watts — every minute. Continue until you can no longer hold the target wattage. Record the highest one-minute power you achieved (your maximal aerobic power, or MAP). Multiply MAP by 0.75 to estimate FTP.
Why 0.75? In trained endurance athletes, FTP typically falls between 72% and 82% of MAP, with 75% as the midpoint. The ramp test works backwards from this relationship: if the test tells us your ceiling, we estimate the fraction of that ceiling you can sustain.
Why it's popular: The ramp test is short — typically 15-25 minutes from start to failure. It requires zero pacing skill. You hold each step until you cannot hold the next one. This removes the single biggest source of error in the 20-minute test and makes the protocol accessible to riders who have never tested before. It is also highly repeatable, which is why Zwift, TrainerRoad, and Wahoo SYSTM have adopted it as their default.
Where it breaks down: The ramp test measures your ceiling, not your floor. It tells you how high you can go, then infers how long you can hold a fraction of that height.
If you are a rider with strong anaerobic capacity — someone who can produce big power for 30-90 seconds — you will push deeper into the ramp than your aerobic engine alone would allow. Your MAP looks impressive. The 75% calculation sets an FTP that your aerobic system cannot sustain. You walk away with a number that makes your sweet spot and threshold intervals feel impossible.
This is not a rare edge case. It is common in riders who came from sports that developed fast-twitch fibres — football, rugby, track and field — and in newer cyclists whose anaerobic system outpaces their still-developing aerobic base.
The reverse is also true. Diesel-type riders — deep aerobic engines, modest anaerobic capacity — often underperform on the ramp test. They reach failure at a lower MAP not because their threshold is low, but because they lack the anaerobic punch to push through the high-wattage stages. The 75% calculation understates their real sustainable power, and their training intervals feel too easy.
Cadence matters on the ramp. As wattage climbs, most riders instinctively drop cadence to produce more force per pedal stroke. This is a muscular strategy, not an aerobic one, and it inflates MAP in the same way that a strong sprint inflates it. If you use the ramp test, try to hold your natural cadence (85-95 rpm) as deep into the test as possible. The moment you start grinding at 70 rpm, you are testing your legs, not your lungs.
The 2×8-Minute Test
Protocol: Warm up thoroughly, then ride two 8-minute maximal efforts separated by 10 minutes of easy recovery. Record the average power of both efforts, take the higher of the two (or the mean, depending on the version), and multiply by 0.90 to estimate FTP.
Why 0.90? Eight minutes is short enough that anaerobic contribution is significant — more so than in a 20-minute effort. The 90% factor compensates for that larger anaerobic subsidy. The protocol was popularised by Carmichael Training Systems (CTS) and has been used extensively with time-crunched athletes who find the 20-minute test mentally daunting.
When it works: The dual-effort structure provides a built-in check against pacing errors. If you go out too hard on the first 8 minutes, you will pay for it on the second. The comparison between the two efforts gives you useful information: if effort one is significantly higher than effort two, you paced badly. If they are within a few watts of each other, you probably found a repeatable intensity.
The test also suits riders who mentally struggle with a 20-minute all-out effort. Breaking the work into two 8-minute blocks with recovery in between makes the effort feel more manageable while still producing a usable number.
Where it falls short: Eight minutes is a short window for estimating an hour-long capacity. The anaerobic contribution is higher than in a 20-minute test, and the 90% conversion carries more uncertainty. Riders with strong anaerobic systems will overshoot for the same reasons they overshoot on a ramp test — the shorter the test, the more anaerobic capacity distorts the result.
The test also introduces a second variable: recovery between efforts. How hard you ride during those 10 minutes, how well you rehydrate, how mentally reset you feel — all of these affect the second effort. Standardise the recovery interval if you want the test to be repeatable.
For riders with the pacing discipline for a 20-minute test, the 8-minute protocol offers marginal convenience at the cost of reduced accuracy. But for riders who consistently blow up in the first half of a 20-minute effort, the shorter format can produce a more honest number than a botched longer one.
The 60-Minute Test
Protocol: Warm up, then ride as hard as you can sustain for 60 minutes. Your average power for that hour is your FTP. No conversion factor needed.
Why it's the gold standard: By definition. FTP is the power you can sustain for approximately one hour. If you ride for one hour at maximal sustainable intensity, the number you produce is the number itself — not an estimate, not a conversion. Every other protocol is trying to approximate what this test measures directly.
Why almost nobody uses it: Pacing a one-hour maximal effort is extraordinarily difficult. Even professional time trialists find it challenging. The margin between "slightly too conservative" and "blew up at 40 minutes" is thin, and most riders who attempt it either pace too conservatively and leave watts on the table, or start too aggressively and fade so badly that the average is lower than a well-paced 20-minute test with the 95% factor would have produced.
The fatigue cost is also significant. A 60-minute maximal effort creates race-day-level physiological stress and demands 48-72 hours of recovery, disrupting the training week in a way that shorter tests do not.
The 60-minute test makes sense for elite time trialists preparing for specific events. For everyone else, a well-executed 20-minute test produces a more reliable number with less disruption.
Which Test Suits Which Rider
There is no universally correct answer. The best test is the one that matches your physiology, your experience, and your ability to execute it consistently. But there are patterns.
If you are new to structured training: Start with the ramp test. It removes pacing as a variable, it is quick, and it will give you a starting number to build from. Expect the result to be approximate — within 5-10 watts of your real FTP in either direction. That is good enough to set initial zones and begin training. You can always retest with a different protocol once you have more experience.
If you are an experienced rider with good pacing instincts: The 20-minute test is the strongest option. You know what threshold feels like. You have the discipline to hold an even effort without starting too fast. The 95% factor has been validated more extensively than any other conversion, and the result will track your fitness changes accurately over time.
If you struggle to pace a 20-minute effort: The 2×8-minute test offers a middle ground. The dual-effort structure is more forgiving of pacing errors, and the psychological burden is lower. Be aware that the 90% conversion carries slightly more uncertainty, and use the comparison between your two efforts as a quality check on the result.
If you have a strong sprint or come from an anaerobic sport: Be cautious with the ramp test. Your MAP will be inflated relative to your sustainable threshold, and the resulting FTP will be too high. The 20-minute or 8-minute test will produce a number more aligned with what you can actually sustain for your training intervals. If your sweet spot sessions consistently feel like threshold sessions, your test overestimated.
If you are a diesel — strong aerobic engine, modest kick: The ramp test may understate your capacity. The 20-minute test captures your strengths better, because sustained high power is exactly what it measures. Some diesel-type riders sustain closer to 96-97% of their 20-minute power for a full hour, so the 95% factor may be slightly aggressive for your profile.
Common Testing Mistakes
Testing errors are systematic. The same mistakes appear across thousands of riders, and most of them are avoidable.
Inadequate warm-up. This is the most common and the most consequential. A cold test — jumping into a 20-minute or ramp effort after five minutes of easy spinning — suppresses the result by 3-8%. Your cardiovascular system needs 15-20 minutes to reach the state where oxygen delivery to working muscles is fully operational.
A proper warm-up includes progressive intensity with two or three short bursts of 30-60 seconds at or above threshold. These bursts are not meant to fatigue you. They prime your neuromuscular system and open up your VO2 kinetics so that your body is ready to produce power from the first pedal stroke of the test.
Going too hard in the first five minutes of a 20-minute test. Adrenaline and the desire to "make it count" push riders above sustainable power in the opening minutes. The physiological cost is non-linear — every watt above your true threshold in the early minutes costs you more than one watt in the late minutes. A rider who starts 15 watts too high for five minutes will typically lose 20-25 watts over the final ten. The net result is a lower average than if they had started conservatively.
The fix: target the first two minutes at a power you know you can hold for 25 minutes. Then settle into a pace that is demanding and just barely sustainable. If you have three minutes left and feel like you could do another ten, you went too conservatively. If you have eight minutes left and feel like you are dying, you went too hard.
Wrong cadence on the ramp test. As described above, dropping cadence late in a ramp test switches the effort from an aerobic measure to a muscular one. The fix: set a minimum cadence threshold — 80 rpm is reasonable — and consider the test over when you drop below it, regardless of whether you could grind out another 30 seconds. Your MAP should reflect what your aerobic system can produce, not what your quads can push at 60 rpm.
Testing in a fatigued state. If you trained hard yesterday, slept poorly, or are deep into a build block without a recovery week, the result will understate your fitness. That lower number then sets your zones too low for weeks. Test after two or three easy days — ideally at the end of a recovery week — so the result reflects capacity, not accumulated fatigue.
Inconsistent conditions. Testing outdoors one time and indoors the next. Different bike, different fan position, different time of day. Each introduces variability that has nothing to do with fitness. Indoor power and outdoor power do not always agree. If you test indoors, always test indoors. Control everything you can so that the only variable is your fitness.
How Often to Test
Every 6-8 weeks, aligned with the end of a training block or the start of a new phase. This is covered in depth in our post on when to test your FTP, but the short version is this: physiological adaptation takes time to manifest in a sustained-power test. Testing every two or three weeks does not show meaningful change — it shows noise from day-to-day variation in sleep, nutrition, glycogen, and stress. That noise creates anxiety, triggers unnecessary zone changes, and adds maximal efforts to your training calendar that produce fatigue without producing information.
If you suspect your FTP has moved mid-block — intervals feel too easy, your heart rate at threshold power has dropped, your RPE at current zones has shifted — adjust zones by 2-3% rather than scheduling a formal retest, and use the interval session builder to recalibrate your workouts around the updated number. Save the full test for a transition point, when you are rested and the result can anchor the next phase of your training.
Once you complete a test, plug the result into the FTP Zone Calculator to set your training zones, and use the TSS Calculator to plan your training load for the block ahead.
Consistency Beats Protocol Choice
This is the point that matters more than everything above it.
If you have been using the ramp test for a year and your results show a steady upward trend, switching to the 20-minute test does not make you fitter. It gives you a different number from a different protocol. You lose continuity, lose the ability to compare against your own history, and introduce new pacing variables. The next six months of data exist in isolation from the previous twelve.
The value of an FTP test is not the absolute number. It is the trend. A rider whose ramp-test FTP has gone from 240 to 260 to 275 over three blocks is demonstrably getting fitter, regardless of whether those numbers would have been 5 watts different on a 20-minute test. The direction matters. The consistency of measurement matters. The protocol is a means to an end.
Use the same test every time. Same conditions — same time of day, same warm-up, same equipment, same indoor or outdoor setting. Record everything: power, heart rate, cadence, RPE, sleep the night before, training load in the preceding week. Over months and years, that record tells a story that a single number never can.
If you are choosing a protocol for the first time, the 20-minute test is the safest default for experienced riders. The ramp test is the safest default for beginners. The 8-minute test is a reasonable fallback for riders who consistently fail to pace a 20-minute effort. The 60-minute test is academically interesting and practically unnecessary for most people.
But once you choose, commit. Run it the same way, every 6-8 weeks, and track the trend. That trend — not the number on any given Tuesday — is the real measure of whether your training is working.