CoachingAnswer

WHICH FTP TEST SHOULD A CYCLIST USE?

By Anthony WalshRoadman CyclingUpdated 26 August 2026

ANSWER

An FTP test estimates threshold power; no protocol is exact for every cyclist. A sustained 45–60-minute effort is close to the FTP construct but demanding, a 20-minute test commonly uses a 95% estimate, and ramp or modelled tests use platform-specific equations. Choose one suitable method, record it with the result, and repeat the same equipment, preparation and conditions before calling a change fitness.

KEY TAKEAWAYS

A test result is method-specific: sustained, 20-minute, ramp and modelled estimates are not automatically interchangeable.
The common 20-minute × 0.95 conversion is an estimate with individual error, not a universal physiological law.
Power source, calibration, cooling, preparation and pacing can change the result independently of fitness.
Use the FTP test calculator only to apply a stated equation; use this guide to choose and interpret the protocol.

WHO THIS IS FOR

IS THIS YOU?

The rider choosing a first test

You need to understand the trade-off between a sustained effort, 20-minute field test, ramp test and modelled estimate.

The rider with conflicting results

Two platforms or protocols produced different values and you need to decide what can be compared honestly.

The self-coached rider resetting zones

You want a repeatable result before changing every power target in the training week.

THE ROADMAN VIEW

The useful FTP test is the one the rider can execute safely, pace appropriately and repeat under comparable conditions. Protocol loyalty does not make an estimate physiologically exact, but it makes the trend more interpretable.

A 20-minute result multiplied by 0.95 is a common field estimate. Published reviews support repeatability while also reporting individual disagreement with laboratory threshold markers. Roadman therefore labels the output as estimated FTP and keeps the equation beside the result.

When a ramp test, a 20-minute test and recent long-duration performance disagree, do not average them into a more scientific-looking number. Check the power source, protocol, cooling, pacing and recent maximal-effort data, then choose the value that best fits the same method and the sessions it will prescribe.

EXPERT EVIDENCE

WHAT THE EXPERTS SAY

Andrew CogganExercise physiologist; originator of the FTP framework
Coggan's published TrainingPeaks guidance treats threshold power as a practical performance estimate and recommends deriving it from appropriate recent efforts rather than presenting one short-test equation as exact for every rider.
British CyclingNational governing body coaching guidance
British Cycling publishes a distinct field protocol and uses the result to organise power-based intensity. The existence of different accepted protocols is a reason to preserve the method name when comparing results.

PRACTICAL APPLICATION

DO THIS WEEK

Choose the method for the rider
Use a sustained effort when pacing skill and a suitable route or race file exist; use a repeatable 20-minute or ramp protocol when practicality matters; use modelled detection only when recent data contain representative hard efforts.
Standardise the comparison
Keep the power source, calibration routine, warm-up, trainer mode or route, cooling and preparation as similar as practical. Record anything that materially changed.
Check the result against riding
Review recent long efforts and the first threshold sessions. A surprising result deserves confirmation; it should not automatically reset every zone.
Send each next task to its owner
Use the calculator to apply a declared conversion, the 20-minute and ramp answers for protocol detail, and the timing guide for when another test is worth the fatigue cost.

COMMON MISTAKES

WHAT CYCLISTS GET WRONG

MISTAKEComparing different protocols as a clean fitness trend.
FIXKeep the test name attached to each number and compare like with like before attributing the difference to training.
MISTAKETreating 95% or 75% as an individual truth.
FIXCall the output an estimate, preserve the platform or protocol equation and cross-check it against representative performance.
MISTAKEIgnoring measurement and environmental changes.
FIXCheck calibration, device consistency, temperature, cooling, terrain and preparation before accepting a large jump or drop.

FAQ

FREQUENTLY ASKED QUESTIONS

What is the best FTP test for cycling?

There is no universally best protocol. A sustained 45–60-minute effort is close to the FTP construct but difficult to pace; a 20-minute test is practical but uses an estimated conversion; ramp and modelled tests are convenient but depend on platform-specific methods. Choose the protocol you can execute and repeat consistently.

How do I calculate FTP from a 20-minute test?

A common method multiplies average 20-minute power by 0.95. That gives an estimated FTP, not a guaranteed one-hour value or laboratory threshold. Use the same warm-up, power source and protocol when tracking change, and keep the 95% method recorded with the result.

Is a ramp test or 20-minute FTP test more accurate?

Neither wins for every rider. Ramp estimates can be influenced by anaerobic contribution and the platform's equation, while a 20-minute result depends heavily on pacing and the individual relationship between 20-minute and longer-duration power. Repeatability for the rider matters more than declaring one universal winner.

Does FTP always equal one-hour power?

No. One-hour power is a useful shorthand, but published research reports substantial variation in how long riders can sustain a measured FTP. Record time-to-exhaustion or the wider power-duration curve when the event requires more context than one threshold estimate.

How often should cyclists repeat an FTP test?

Retest when the current value appears stale, after a meaningful training block, or when repeatable workouts show the zones no longer fit. No fixed calendar interval suits every rider. The timing guide covers fatigue, block placement and alternatives to formal testing.

Why did two FTP tests give different results?

Different equations, pacing demands, anaerobic contribution, power meters, calibration, cooling, fatigue and fuelling can all change the result. First identify which variables changed; then repeat one method under comparable conditions before interpreting the gap as fitness.