FTP Is Dead? Why Critical Power Is Replacing FTP in Professional Cycling

The Problem With FTP

Functional Threshold Power is defined as the highest power you can sustain for approximately one hour. In practice, almost nobody tests it with an actual one-hour effort. Instead, we estimate it from 20-minute tests, ramp tests, or 8-minute protocols, each with its own correction factor.

The fundamental issue: FTP is a single point on your power-duration curve. It tells you nothing about your 5-second sprint, your 5-minute VO2max capacity, or your anaerobic reserve above threshold. Two riders with identical FTPs can have wildly different strengths and training needs.

Worse, the "one hour" definition is physiologically imprecise. Research by Karsten et al. (2021) showed that the duration an athlete can sustain FTP ranges from 30 to 70 minutes depending on their physiology. A sprinter-type might only hold their "threshold" for 35 minutes, while a diesel endurance rider might sustain it for over an hour.

What Critical Power Actually Is

Critical Power (CP) is the boundary between sustainable and unsustainable exercise. Below CP, your body can achieve a metabolic steady state — lactate stabilizes, oxygen consumption plateaus, and you can theoretically continue indefinitely. Above CP, physiological reserves are depleted at a predictable rate until exhaustion.

The CP model has two parameters: CP itself (the power at the boundary) and W' (pronounced "W prime"), which represents your total work capacity above CP. Think of W' as a finite battery of anaerobic energy — every watt above CP drains it, and recovery below CP recharges it.

Together, CP and W' describe your entire power-duration curve. At any duration, your maximum sustainable power is:

Power = CP + (W' / Time)

This single equation predicts your 3-minute, 5-minute, 12-minute, 20-minute, and 60-minute power. It explains why you fade, when you will fade, and how hard you can go for any duration.

CP vs FTP: A Side-by-Side Comparison

W' — Your Anaerobic Battery

W' is measured in kilojoules (kJ) and represents the total amount of work you can do above CP before exhaustion. A typical value for a trained cyclist is 15–25 kJ. Sprinters tend to have higher W' values (25–35 kJ), while endurance riders often have lower W' (10–18 kJ).

What makes W' powerful is that it depletes and recovers during a ride. Every second above CP drains your W' battery. Every second below CP recharges it (following a time constant of roughly 300–600 seconds). This creates a real-time model of how much you have left.

This is W'bal — the W' balance at any point during a ride. When W'bal hits zero, you are done. You cannot sustain power above CP. This is the scientific explanation for "blowing up" in a race.

How CP Is Calculated

The classical method requires two or more all-out efforts at different durations. The standard protocol is a 3-minute and a 12-minute all-out effort, separated by 30 minutes of rest. The math fits a hyperbolic curve to your power-time data, yielding both CP and W'.

Modern software takes a different approach. Platforms like WKO5, Xert, and Golden Cheetah extract CP and W' from your accumulated ride data. Every hard effort you do — whether a sprint in a group ride, a hill climb, or a race — contributes data points. The model updates continuously without dedicated testing.

This is a major practical advantage. Many riders never perform a formal test but accumulate dozens of hard efforts over weeks of riding that collectively define their power-duration relationship with high precision.

The 3-Minute All-Out Test

The Vanhatalo et al. (2007) study validated a single 3-minute all-out test for estimating CP. After an easy warmup, you ride as hard as possible for 3 minutes. CP is estimated as the average power of the final 30 seconds, and W' is the work done above that average.

This test is brutally hard but effective. The key requirement is truly maximal effort from the start — if you pace it, the results are invalid.

When CP and FTP Diverge

For most trained cyclists, CP and FTP are within 3–8% of each other, with CP typically being slightly lower. But the gap varies by rider type, and this is where the single-number limitation of FTP breaks down.

Sprinter types tend to have high W' relative to CP. Their FTP (estimated from a 20-minute test) may be artificially inflated because their anaerobic capacity contributes significantly during a 20-minute effort. They end up training in zones that are too high.

Diesel/endurance types tend to have low W' relative to CP. Their 20-minute test underestimates their true sustainable power because they have little anaerobic contribution. They end up training in zones that are too low.

The CP model handles both rider types correctly because it separates the aerobic (CP) and anaerobic (W') components. No correction factor needed.

Who Is Already Using CP?

The shift is already happening in the tools serious cyclists use:

• WKO5 by TrainingPeaks uses a modified CP model (mFTP/mCP) as its core metric, replacing the manual FTP entry of earlier versions
• Xert built its entire platform around a CP-like model with Threshold Power, High Intensity Energy (W' equivalent), and Peak Power
• Golden Cheetah has supported CP/W' modeling since 2015 and uses it for interval prescription
• intervals.icu calculates both CP and W' from your ride data alongside traditional FTP
• Garmin and Wahoo still use FTP on their head units, but Garmin now shows power curve data that is essentially a simplified CP visualization

Why FTP Still Persists

Despite its limitations, FTP is not going away tomorrow, and there are legitimate reasons why:

Simplicity: One number is easier to understand, communicate, and compare than two. "My FTP is 280" is immediately meaningful. "My CP is 270 and W' is 18 kJ" requires explanation.

Infrastructure: Every training platform, coach, and training plan in cycling uses FTP. Power zones, TSS, IF, workout prescriptions — all built on FTP. Switching to CP requires rebuilding the entire training metrics ecosystem.

Good enough: For the majority of recreational cyclists, FTP works well enough for setting zones and tracking progress. The divergence between FTP and CP matters most at the extremes — pure sprinters and pure endurance athletes.

What This Means for Your Training

If you are currently training with FTP-based zones and seeing progress, there is no urgent reason to switch. FTP will continue to work for setting training intensities and tracking fitness trends.

However, understanding CP and W' adds valuable context:

• If your zones feel too hard, you may be a sprinter type whose FTP is inflated by high W'. A CP-based threshold would be lower and more accurate.
• If your zones feel too easy, you may be an endurance type whose FTP is deflated by low W'. CP would set a higher threshold.
• For race pacing, W'bal tracking tells you exactly how much you have left for attacks, surges, and the final sprint — something FTP alone cannot provide.

The Future: CP Goes Mainstream

The trajectory is clear. As more platforms integrate CP/W' modeling and riders become comfortable with two-parameter fitness profiles, FTP will gradually recede from its dominant position.

The most likely outcome is not that FTP disappears, but that it becomes one output of a CP model rather than a standalone metric. Your "FTP" becomes the power you can hold for one hour as predicted by CP and W' — a derived value rather than a tested one.

For now, the best approach is to understand both models, train with whichever your platform supports, and pay attention to your entire power-duration curve rather than fixating on a single number.

Further Reading

The CP model was first described by Monod and Scherrer (1965) and has been validated across hundreds of studies. Key papers include Morton (1996) on the critical power concept, Jones et al. (2010) on the physiological meaning of CP, and Vanhatalo et al. (2007) on the 3-minute all-out test protocol.