FTP (Functional Threshold Power) and CP (Critical Power) both answer the question "how hard can I go for a long time?" But they answer it differently - and using the right one matters for your training.
FTP: the practical standard
FTP, popularised by Dr. Andrew Coggan, is a single number: the highest power you can sustain for approximately 60 minutes. In practice, most people estimate it from a 20-minute test (95% of 20-min power) or a ramp test (75% of peak power).
FTP's strength is simplicity. One number. Multiply by percentages to get zones. Easy to track over time. The whole cycling world uses it.
FTP's weakness: it assumes a single threshold. Your sustainable power for 30 minutes, 60 minutes, and 90 minutes are different - FTP flattens them into one number. For riders with an anaerobic bias (strong short efforts, weaker sustained power), FTP overestimates sustainable power. For diesel engines (weak sprints, strong endurance), FTP underestimates it.
Critical Power: the mathematical model
Critical Power comes from a different tradition - the critical power model from exercise physiology. It uses two parameters:
- CP (Critical Power): the power you can sustain indefinitely in theory (in practice, ~30–60 minutes)
- W' (W-prime): your anaerobic work capacity - total work above CP you can do before exhaustion (measured in kilojoules)
These are estimated from two maximal efforts at different durations - typically 3 minutes and 12 minutes. CP ≈ your 12-minute best power (roughly). W' ≈ the extra work your 3-minute power produced above CP, scaled up.
The CP model is more accurate for predicting performance across time scales. Your 5-minute power, 20-minute power, and 60-minute power can all be estimated from CP + W'. FTP can't do that.
When to use which
Use FTP when:
- You're following a training plan that uses Coggan zones (TrainingPeaks, TrainerRoad, Zwift)
- You want a single number to track and share with your coach
- You don't have reliable maximal efforts at two different durations (required for CP)
Use CP when:
- You want to predict performance at durations you haven't tested
- You're analysing your power curve to find strengths and weaknesses
- You're self-coached and want deeper understanding of your physiology
The practical answer:
Use both. FTP for day-to-day zone-based training. CP to understand your power profile and track your W' (anaerobic capacity) - which FTP can't tell you anything about. Baseline computes both automatically from your ride data.
What the gap between FTP and CP actually means for your training
When your CP is higher than your FTP (the most common case for athletes who test their 20-minute power), your sustainable aerobic power is actually better than your FTP suggests. This gap typically ranges from 5–15 watts for amateur cyclists and up to 25 watts for elite riders with large aerobic engines. If CP is 260W and FTP is 245W, your zone 4 threshold work at 245W is systematically undershooting what your physiology can actually sustain. You are training slightly below your true threshold.
The opposite case - CP lower than FTP - is rarer but more dangerous. It signals that your FTP test overestimated your sustainable power, usually because your 20-minute effort included a significant anaerobic contribution. Your zones are set too high. Every threshold interval becomes a VO2max effort. This is the classic "I hit my FTP zones in January but by March I'm burnt out" scenario. If your CP is 10+ watts below your FTP, recalculate your zones from CP and retest after 4 weeks of zone-based training.
The W' number carries its own training signal. A high W' relative to your body weight (> 250 J/kg) indicates a strong anaerobic profile - you can repeat short high-power efforts and recover quickly. These athletes benefit most from explosive interval work (30/30s, 40/20s) and track work. A low W' (< 180 J/kg) suggests a diesel profile: prioritising long threshold intervals and tempo blocks will deliver more than sprint training. W' also drops with accumulated fatigue, making it an excellent daily readiness indicator. If your usual W' of 18 kJ drops to 12 kJ on a given day, your anaerobic battery is depleted — skip the sprint session and do an easy ride instead.
Common mistakes
1. Treating FTP as a fixed number. Many athletes set their FTP once and never update it for months on end. FTP drifts - it increases with fitness and decreases with time off or reduced training load. Using a stale FTP means every zone is wrong. Intervals that should be threshold become too easy or impossibly hard. Retest every 4-6 weeks during a build phase and every 8-12 weeks during maintenance. Your power zones are only as accurate as your FTP.
2. Ignoring the full power duration curve. FTP gives you one data point - your approximate 60-minute power - but your performance at 1 minute, 5 minutes, and 20 minutes tells a much richer story. Two athletes can both have a 250W FTP yet have completely different power profiles. One might excel at short anaerobic efforts while the other is a diesel engine. The FTP number alone hides these differences, which means your training prescription should be different even though your zones are the same.
3. Confusing CP with FTP. Critical Power and FTP are related but not interchangeable. CP is the asymptote of the power-duration curve and comes paired with W' (anaerobic work capacity). FTP is a practical estimate based on a single test protocol. Using CP values directly in an FTP-based zone system produces incorrect intensities, especially for shorter efforts where W' plays a significant role.
How Baseline handles this
Baseline computes both FTP and Critical Power automatically from your existing ride data - no special test required. The power profile view shows your best efforts across all durations from 5 seconds to 4 hours, plotted against your calculated CP curve. The dashboard updates your FTP whenever you log a new best effort, so your zones stay current without manual retesting.
Further reading
How to choose your test protocol
The protocol you use to derive your FTP or CP determines the quality of the number, and protocol selection should match your physiology. Athletes with strong anaerobic profiles (sprinters, track cyclists, criterium racers) typically get inflated FTP estimates from a 20-minute test because their anaerobic contribution to a 20-minute all-out effort is significant. For these riders, the critical power model with 3-minute and 12-minute maximal efforts produces a more accurate threshold estimate by factoring out the anaerobic contribution via W'.
Diesel-type endurance athletes (time trialists, randonneurs, triathletes) with relatively weak anaerobic capacity actually get more accurate FTP estimates from the 20-minute test because less anaerobic contribution means the 95% correction factor is closer to their true threshold. Their 20-minute power and 60-minute power are close together, so a single test works well. For these riders, the key is not the method but the execution: a proper 5-minute blowout effort before the test is essential to drain whatever anaerobic capacity exists before the measured effort begins.
The simplest heuristic: if your power drops by more than 20% between a 1-minute maximal effort and a 20-minute maximal effort, your anaerobic contribution to the 20-minute test is high and you should use the CP model. If the drop is less than 15%, the 20-minute test is reliable for you. Most well-trained amateur cyclists fall in the middle of this range and benefit from tracking both numbers over time.