Watt Matters is a blog about training and racing with power and other related musings.

Aside from writing items here on occasions, I also provide cycling performance improvement services via coaching, aerodynamics testing and host a cycling tour.

If you'd like to ask me a question or suggest a topic you'd like my take on, then just head to Ask Alex and let me know!



Functional Threshold Power - A Definition

This is a long post about the definition of Functional Threshold Power (FTP). 

I’m not going to be discussing the utility of FTP, nor how to estimate it or even how to train to improve it. While each are worthy topics, they are not what today’s inforant is about.

What I will do is provide some thoughts about what threshold power represents (and a bit of what it doesn’t).

Seriously, how hard can it be? We all know what FTP is, right?

That said there is a more nuanced understanding that’s seemingly been lost over the years, partly due to the widespread adoption and success of the concept of FTP but also resulting from some unintentional as well as deliberately misleading information put out there. So I thought I’d revisit the discussion.

Misunderstandings and misrepresentations about FTP appear all the time in various online forums, media articles, training videos, cycling business websites and social media. And sometimes in published scientific literature. I can’t vouch for post-ride coffee shop chats around the world but I’m guessing the same mistakes have been repeated and for the most part the misunderstanding is just that, a misunderstanding.

Does it matter? 

We don’t each need to understand the amazing complexity of human biology in order to know how to improve our athletic performance and as such minor misunderstandings about the definition of threshold power are not so grave as to make much difference in how we go about our training.

Sure, having a grasp of the underlying physiological fundamentals can provide useful insight but for the most part we need only concern ourselves with macro-world factors and evidence-based principles for performance improvement.

It does matter that we keep our definitions in-tact and relevant else they become so distorted as to no longer be useful, not based upon reality and/or can result in poor advice or decisions about our training.

Also, such a useful and sound concept as Functional Threshold Power (FTP) is occasionally deliberately misrepresented or unfairly attacked.


I did a little background research to see if I could nail down the earliest use of the term Functional Threshold Power. From what I can see it entered the power training lexicon sometime in early 2002, with one of the earliest examples of the phrase being used in this Topica Wattage forum post by Andy Coggan. The original and now defunct Topica Wattage forum is not always up and running, so this is the equivalent reference in the Google Groups Wattage forum - but to view it you'll need to be a member, or apply to join.

There were seeds of the term appearing in the year before this and it's quite probable Andy used it before then but this is near enough to its genesis. It certainly became a frequently used term from that point on. For those of us who've been engaging with the topic of training and racing with a power for the best part of the past two decades, the term is iconic.


Question: What is FTP?

Typical answer: The power you can maximally sustain for about an hour. 


Well, sort of. There’s a little more to it than that.

Hunter Allen, co-author with Andy Coggan of the book Training and Racing with a Power Meter, wrote this on his blog five years ago:

So what exactly is FTP?

Do you want the short answer or the long answer?

In simplest terms, your functional threshold power, or FTP, is the maximum power you can maintain through an hour’s effort without fatiguing.

But it’s actually much more complicated.

The word “threshold” has become synonymous with the word “confusion” for many athletes.

Before going further, below is a link to an article by Andy Coggan to explain FTP. Straight from the exercise physiologist's mouth (although it's entirely possible the host has modified it!):

What is Threshold Power?

Let’s go back to the original definition by Andy Coggan, since he came up with the concept of Functional Threshold Power (as it appears in Training and Racing with a Power Meter, page 44 in first edition and page 41 in the second edition):

"FTP is the highest power that a rider can maintain in a quasi-steady state without fatiguing for approximately one hour. When power exceeds FTP, fatigue will occur much sooner, whereas power just below FTP can be maintained considerably longer".

OK, so what’s the problem? 

Nothing, it’s a perfectly fine statement; it aligns nicely with physiological reality, provides a useful indicator as to how hard such an intensity is as well as how to determine it, and it works really well in practice. What’s not to like?

A key nuance that is perhaps missed is that "threshold" is a physiological phenomenon, and FTP expresses this intrinsic physiological/metabolic capability in power output terms.

As part of various attacks on the definition of FTP, some focus on the “about an hour” part, some the “quasi-steady state” part. Some want it tied to something else they are familiar with such as lab tests. Often what they seek is a false sense of precision, to describe the phenomenon in unrealistic terms, or to ascribe an unnecessary evaluation process. Then there are those who criticise FTP not because of definition issues but because of the (sometimes not recommended) manner in which it is estimated.

Expressing threshold power in terms of a precise duration or to suggest the relationship between energy demand and physiological responses at threshold is not representative of a quasi-steady state system would be to ignore physiological reality.

A precise duration at which all individuals can sustain power at threshold just doesn’t exist (factors influencing fatigue are many and variable), and quasi-steady state is exactly the right terminology for what’s going on, as we shall see a bit later on. 

As to being tied to some lab method, well that kind of misses the point of it being functional. Labs are great for research and controlling things but what’s of most use is measuring how we actually perform out on the bike, understanding our functional performance factors, how they relate to our goals and developing insight into how to train them.

So let’s examine the terms in this definition to provide some additional context.
I’ll explore in turn:
•    Threshold
•    About an hour
•    Quasi-steady state
•    Power
•    Functional


Threshold is a conceptual expression of an intrinsic physiological capability, i.e. the maximal exercise intensity one can sustain for a long duration.

This is often further refined to mean the maximal exercise intensity one can sustain so that various (relevant) physiological responses firstly attain and can maintain a quasi-steady state, and/or is just below an intensity at which non-sustainable physiological/metabolic processes begin to manifest themselves in a significant manner.

Naturally these physiological responses will also attain/maintain a quasi-steady state at exercise intensities below threshold. Sub-threshold intensity can be sustained for quite long periods (up to a lifetime in fact but e.g. we maintain a physiological steady state when we sleep, sit, walk, as well as when we cycle steadily at moderate effort level).

There are of course a large handful of ways in which physiological responses are indicated or measured, e.g. via invasive methods (e.g. taking blood samples for measurement of blood lactate or plasma lactate levels or to assess hormone response, even insertion of needles to perform an electromyography of working muscles ) or slightly less invasive means (e.g. gas exchange analysis, breathing rates, infra-red muscle tissue oxygenation monitors). Each are proxy markers for the underlying physiological processes.

A myriad of research testing protocols have evolved over the years as well as analysis techniques. Each has their strengths and weaknesses and repeat-ability varies.
The plethora of such protocols goes some way to explaining why there is some confusion over the concept of threshold. By and large though most are describing or seek to indicate the same basic phenomenon, and typically the intensities at which each occurs (as defined by its own method) are all correlated.

So what’s going on?

Doing work involves lots of complex biological wizardry and how much and how hard we can work boils down to providing enough Adenosine Triphosphate (ATP) to meet the body’s energy demand. ATP is a special molecule involved in biochemical reactions to provide for our energy needs – it’s the energy currency transfer mechanism of almost all life. It’s produced primarily within the mitochondria inside our cells and we are constantly generating and recycling ATP via various metabolic processes.

Our aerobic (with oxygen) metabolism can supply our body with energy all-but indefinitely; however the rate at which we can supply this energy is limited. While ever we operate in a state such that the energy demand is maintained at or below this rate limit, then our physiological responses will attain and then maintain a quasi-steady state.

However if we require energy at a rate in excess of this limit, our bodies respond by supplying and generating ATP via non-sustainable capacity-limited energy supply sources (e.g. via anaerobic metabolism). So while this supplemental energy supply is available at a relatively rapid rate, there is only a limited energy capacity available to use before it runs out and needs to be recharged.

One consequence of our bodies drawing more heavily upon these supplemental metabolic processes is they result in more rapid changes in physiological markers and these no longer maintain a quasi-steady state, e.g. blood lactate utilisation/clearance can no longer keep up with production and so blood lactate levels rise continually. 

Here's one example of what I mean. Research conducted on 26 men.

Methodological aspects of maximal lactate steady state—implications for performance testing

A chart from the study showing test subject’s blood lactate levels while riding:


We can see that in both tests during the first 5- to 10-minutes the blood lactate level rises rapidly from a baseline level.

When cycling at a power associated with MLSS, the blood lactate level settles into a quasi-plateau after approximately 10-minutes.

At a power output 6% higher than MLSS the blood lactate level never plateaus but rather continues to rise at a linear rate after the 10-minute mark.

Another consequence of drawing upon these non-sustainable energy sources is that ultimately fatigue will occur much sooner. Indeed many of the subjects in the above study could not even complete the 30-minute trial at the 106% level. 

Hence the term "Threshold". There are several scientific papers which explore various threshold indicators. Here is one reasonable example examining the literature in which they identified 25 different lactate threshold concepts alone.

Lactate Threshold Concepts. How Valid are They?

About an hour

The inclusion in the definition of FTP of an approximate duration at which one can sustain this threshold intensity has of course has been misrepresented over the years and sometimes morphed in "exactly 60-minutes" or “95% of 20-minute power”.

There is individual variability in the duration threshold power can be sustained. It’s a physiological reality and is the case no matter which definition of threshold you refer to.
The maximal power one can sustain for exactly one-hour would of course be a very good estimate of FTP, but it is not FTP itself. FTP has never been defined that way, and for good reason.

95% of your 20-min power is also just a method of estimation and not a definition. In this case it’s not a method I’d recommend due to large individual variability in such ratios.

Ways of estimating an intrinsic physiological phenomenon are not the same as the phenomenon itself.

FTP is partly a victim of its own success. 

A quick Google search of the phrase "functional threshold power" will yield no end of articles, videos, business enterprises and so on many of which repeat the same misunderstandings. Most do understand its value, utility and ways to estimate it, which in practical terms is what matters.

Does Threshold have a duration?

All of the various other "thresholds" described in the scientific literature share an important feature in common with FTP - none of them specify a precise duration
Rather they refer to maximal intensity sustainable:

  • for "a long time" and/or

  • at which a quasi-steady state in physiological responses is still observed, and/or

  • before non-sustainable physiological responses begin to manifest themselves in a significant manner.

It is rather amusing to read criticism of FTP not being tied to a precise duration. No definition of threshold is.

That said, physiological responses to a given energy demand not only change with intensity but also with time (and are also influenced by other factors, fitness of course but also glycogen levels, fatigue and so on). Exercise physiologists do place practical limits on the rate of such time induced change before it’s no longer considered a quasi-plateau response. Be they blood lactate levels, tissue oxygenation level or other proxy markers.

As it turns out, when time to exhaustion tests are performed at intensities defined by well-established laboratory measured threshold markers such as Maximal Lactate Steady State, the duration cyclists can sustain such an intensity is typically in the 40-minute to 70-minute range. Sound familiar?

As a rule of thumb, better trained cyclists can sustain power at threshold for longer than the un- or under-trained. This is exactly what we see for instance when we examine results from time to exhaustion (TTE) tests for cyclists riding at their Maximal Lactate Steady State power:
Fifteen moderately trained men (TTE: 37.7 ± 8.9 min)
Fourteen trained male cyclists (TTE: 54.7 ± 10.9 min)
Eleven male trained subjects (TTE: 55.0 ± 8.5 min)

And in this study we see that power at lactate threshold, 1-hour ergometer power and 40-km time trial performance were strongly correlated:
Physiological and biomechanical factors associated with elite endurance cycling performance.

Quasi-steady state

There are two levels to understanding the term quasi-steady state.

The simplest is its broad definition:

A situation that is changing slowly enough that it can be considered to be constant.

That’s pretty straightforward and we’ve already seen an example of that in the blood lactate chart above, but there’s a bit more to it. When we consider quasi-steady state in systems, in our case the human biological system, it has additional meaning.

This neat explanation provides some insight into what this phrase means in a system sense:

“… in general the quasi-steady state assumption is used when one part of the system reacts much more quickly (read has a shorter characteristic timescale/equilibrates faster) than another part. When one part of the system equilibrates faster than another part we can say that this part of the system is basically in steady state with the slower moving system (approximately). We can then simplify the time dependence of the faster reacting part of the system to the slower equilibrating part of the system which reduces the number of variables we need to solve for.”

In the case of human biological systems, our power output, neuromuscular function and biochemical processes (especially with the aid of enzymes) can change/respond over quite short time scales (seconds or fractions of a second) but many of our physiological responses operate on a far slower time course (minutes).

So despite variability in these short term processes, provided the duration of fluctuations are not long and they bounce around a sustainable mean value enabling the slower reacting physiological responses to attain a steady state, then the entire system can be considered to be in a quasi-steady state.

It’s akin to looking at a trace of power output smoothed to show a rolling 30-seconds average power compared with the raw data showing the second to second variability in power output (or indeed the highly variable power output during each pedal stroke). From a metabolic perspective, it’s the longer duration average that matters for whether or not a quasi-steady state can be considered to apply.


FTP expresses this threshold intensity in terms of mechanical power output (at the cranks of the bicycle). So that’s the P in FTP. No more, no less.


The "Functional" part of FTP is simply to emphasise its purpose is to be a practically useful measure of an intrinsic physiological capability.

The functionality of FTP is further demonstrated by the ability of just about anyone to measure their power output on a bicycle, to conduct tests or examine race efforts as needed, whereas the various physiological responses typically measured in labs are somewhat less practical for the regular athlete to measure and are also somewhat invasive. We can gather power data from every moment of every ride. Lab tests, if people bother with them, are infrequent snapshots.

And then there is the question of which lab test? As I mentioned earlier, there are a multitude of different lactate test protocols and it is confusing.

Putting aside academic interest, power output where and when we ride is of far more interest than what occurs in a lab. And FTP is about our actual performance capability, out on the road, track or wherever you ride or race.

Power output has the added advantage of being an integral of all the underlying physiological factors (e.g. efficiency, VO2max, fractional utilisation of VO2max and so on). So rather than be bogged down with physiological, biological and metabolic complexity, we can express this intrinsic physiological capability in one simple and useful term - FTP.

There are really only a small handful of intrinsic physiological characteristics that differentiate cycling performance capability with threshold power being one, while others define our non-sustainable supra-threshold power abilities (and in a sense the balance of these characteristics determines how we race).

In endurance cycling, threshold power is the most important of these physiological characteristics. Naturally there are many non-physiological factors which are also important for performance.

There are also various ways in which one can arrive at an estimate of FTP, some more reliable than others (hint: nothing beats actual performance). The reliable methods haven't changed much over the years, except perhaps one area which has evolved is the use of more sophisticated power-duration models. That's a discussion for another day.

In practice, as athletes and coaches we really don't really need to know all the underlying nuances - we are interested in things which help us to measure and improve our performance. The core success of FTP (aside from its basis as a real physiological phenomenon) stems from Functional part, and somewhat ironically this is probably a reason why there is confusion or limited understanding of what it actually is.

FTP provides excellent indicator of metabolic fitness and it's relatively easy to estimate with various methods that can be adapted to suit the individual athlete. What's not to like?

Beyond that, does it matter? Probably not a whole lot.

Functional Threshold Power in the scientific literature

Here’s a list of some published research about FTP in order of publication date. I quote from the full text the definition of FTP they use so we can see how well or not they have represented it:

February 2012
Comparison of a field-based test to estimate functional threshold power and power output at lactate threshold. 

Functional threshold power (FTP) has been defined as the highest average power output that can be maintained for 1 hour

June 2014
Validity of using functional threshold power and intermittent power to predict crosscountry mountain bike race outcome 

With the increasing availability of on-the-bike personal power measuring devices, cycling coaches developed functional threshold power (FTP) as a field test to estimate the LT (Allen & Coggan, 2006). FTP can easily be measured during a twenty-minute maximal-power time-trial and tracked throughout a training program with training intensity zones and athlete ability subsequently estimated (Allen & Coggan, 2006).

March 2017
A field-based cycling test to assess predictors of endurance performance and establishing training zones 

The FTP was defined as the highest average power that a rider can maintain for 60 min and serves as an individual’s estimated power output at lactate threshold (defined as 1 mmol·L-1 increase above baseline)

June 2017

Relation between Critical Power and Functional Threshold Power 

FTP has been defined as the highest average power output (PO) that can be maintained for 60 min. CP testing requires multiple exhaustive trials which might explain the apparent gap between research and real-world cycling whilst FTP can be determined from a single maximal 20 min effort.

September 2017
Cycling Power Outputs Predict Functional Threshold Power And Maximum Oxygen Uptake 

Functional threshold power (FTP) is the maximum average power sustained over a 1-hour period, which is estimated through 8- or 20-min self-paced cycling protocols by subtracting 10% or 5% of the average power achieved after each respective cycling test

May 2018
Is the Functional Threshold Power (FTP) a Valid Surrogate of the Lactate Threshold?

One of the most commonly used field-test methods for the estimation of the LT is the functional threshold power (FTP), which has been defined as the highest power that can be maintained in a semi-steady state.

In practical terms, the FTP corresponds to the highest mean power that athletes can sustain for around 1-h under competitive conditions (e.g., highly motivated, well rested, appropriate nutrition)

May 2018
Functional Threshold Power in Cyclists: Validity of the Concept and Physiological Responses 

October 2018
Correction: Functional Threshold Power in Cyclists: Validity of the Concept and Physiological Responses

Functional threshold power (FTP), defined as the highest PO a cyclist can maintain in a quasi-steady state for approximately 60min without the onset of fatigue.

May 2018

The Reliability of 4-min and 20-min Time Trials and Their Relationships to Functional Threshold Power in Trained Cyclists.

“the highest power that a [cyclist] can maintain … for ~1 hour …” – was termed Functional Threshold Power (FTP) by Allen and Coggan

Updated since initial post:

August 2018
Influence of lateral pedal translation on muscle recruitment and kinematics in cyclists.

Participants were first asked to complete a test of their functional threshold power.14 Functional threshold power (FTP) is the maximum cycling output power that a cyclist can maintain for an hour period. The test consisted of a warmup period, an 8 min period of maximum effort, and a cooldown.15 Furthermore, rather than determining a precise FTP for a subject, the intent of the FTP test was to establish an effort level for their pedal testing.

October 2018
Low energy availability assessed by a sport-specific questionnaire and clinical interview indicative of bone health, endocrine profile and cycling performance in competitive male cyclists

In road cycling, a high power-to-weight ratio, expressed as 60 min functional threshold power (FTP) watts/kg confers a performance advantage. Cyclists quantify FTP according to BC protocol (or equivalent), typically early in the cycle season to set their training zones. This self-reported metric was recorded in the sport specific questionnaire, together with BC race category (which can be verified on the BC website).

January 2019
Is the Functional Threshold Power Interchangeable With the Maximal Lactate Steady State in Trained Cyclists?

Functional Threshold Power (FTP), determined as 95% of the average power during a 20-minute time-trial test

March 2019
Intermittent Pneumatic Compression Boot Use Elevates Blood Lactate During Subsequent Exercise

a functional threshold power (FTP) test was performed. tests are a commonly utilized, maximal effort field test and are a valid prediction tool for cycling performance (13). FTP is calculated as 95% of the mean wattage over the 20 minute testing period.

April 2019
Road cycle TT performance: Relationship to the power-duration model and association with FTP.

the FTP test started with 3 min of baseline pedalling at < 90% GET at preferred cadence, followed by a 20-min maximal, self-paced TT. The FTP was defined as 95% of the mean PO achieved during the 20-min TT

That's it, inforant over.

Cover image credit:


Alpe d'Huez: TDF Fastest Ascent Times 1982-2018

Alpe d'Huez: TDF Fastest Ascent Times 1982-2018

The Confusion of Cycling Advocacy in Australia

The Confusion of Cycling Advocacy in Australia