Endurance 101: Zone 4 - Functional Threshold Power (FTP) Made Simple

What is FTP?

"FTP" is an acronym for "Functional Threshold Power". It continues to be one of the most misunderstood and poorly communicated concepts within cycling. In order to understand what FTP is, we're going to work "backwards", deconstructing it from the most common mis-understanding back, unpeeling one layer of the puzzle at a time.

FTP is the the maximum power that I can hold for 20 minutes, right?

No. There's a lot "right" about this statement, but also a lot wrong. The thing that is most correct about the statement "My FTP is my 20 minute power" is that it recognises that FTP is a field measure of power-that is, a functional measurement of power output.

By functional we here mean all of the most common dictionary definitions - it "has a special. purpose" (is tied to physiological parameters of significance) and also that it is "designed to be practical and useful" (practical in that it can be measured or estimated in the field without specialised lab equipment, and also useful in that it has significance in both describing something about our current training state, and also as an input when prescribing effective training for us to follow).

But it's not equal to - or tightly predictable from - 20 minute maximal power.

But I do a 20 minute test to determine my FTP!

Not if you are measuring your FTP properly you don't ;-). Your twenty minute power does not of necessity correlate tightly with your FTP.

This misunderstanding arises because one of the original test protocols for estimating FTP featured a 20-minute effort. But that 20-minute effort was intended to form part of a very specific test protocol, in which the 20-minute effort was preceded by a 5-minute interval performed all-out at maximal power ("VO2 max"). The rationale for this was to draw down a portion of what is now referred to as "W prime" or "FRC" (functional reserve capacity) in order to minimize anaerobic power contribution to the succeeding 20-minute effort. It was also secondarily designed to "prime" the aerobic system via activation of the VO2max "slow component". Without this 5-minute preceding effort, and for both the reasons above, a 20-minute power value may relate much less precisely to threshold aerobic power. In the original test protocol, the resultant 20-minute power figure (average, not normalised) was also then multiplied by a factor of 0.95 (95%) in order to estimate threshold power.

Since this test protocol was established it has has been diluted and miscommunicated until it has become widely referred to as "ride for 20 minutes to find your FTP" by many channels. Given this, along with other reasons we shall shortly see, Zwift most likely over-estimates many riders' FTPs.

But before we critique that particular test prescription-or any others-lets try and understand what, if anything, is special about FTP power.

Why test at 20 minutes? What makes FTP power important at all?

Some sports are constrained to fairly primitive measures of performance, such as "pace" (distance covered in a given time). The limitation here is that, unless conditions are exactly equivalent, pace may vary.

In cycling, we have "power meters" either as stand-alone or in the turbo trainers that enable us to record our power output in watts. A watt is a unit of power or "radiant flux" that is equivalent to 1 joule per second. This is a unit that defines the rate of energy transfer.

Watts are not specific to cycling-a 100 kg man who climbs a 3 metre ladder in 5 seconds, for instance, is performing work at a rate of approximately 600W. Cycling is, however unique among popular sports in that work rate can be readily and reasonably accurately measured.

Since watts are a measure of energy per second, the use of a power meter allows us to capture power outputs for different lengths of time.

During a ride, there is an inverse relationship between power and duration - we can sprint for 5 seconds (high power, short duration) or walk all day, at low power and high duration. We cannot, however, sprint all day. So, training takes in these two inputs in some combination-and effective training finds the balance between them. We can plot power output across the differing durations on a graph, commonly referred to as a "powerduration curve" or PDC, and one to which you may be familiar via various online training platforms. These graphs may show an additional estimated FTP at a given point on this curve, typically at or around the hour mark.

I know this, I'm not one of those "20 minute" suckers, my FTP is my one hour power, right?

No.

In fact, FTP does not tie precisely to any one, single duration. That is, FTP is a metric of power with a variable duration. The length of time you can sustain your FTP is called your "TTE" (Time To Exhaustion) and this length of time, or TTE is trainable, just like your power at FTP. A typical TTE for FTP might fall in a range of 40-70 minutes. This will be an important point, later. But first, if FTP does not map to your maximum sustainable power over an hour, what does it map to exactly? Why is the TTE for FTP variable?

Understanding why this doesn't always relate to maximum power across precisely one hour requires a little deeper delving into the science-if this all sounds a bit dry then please stick with me, because fully understanding the science means you'll be able to properly optimise your FTP training.

FTP is a significant point on our PDC because of the underlying physiological state that it corresponds to within our body.

In our previous post "Zone2 101 - Everything you ever wanted to know about Zone2" we outlined what mitochondria were and how the measurement of blood lactate can be used to tell us something about our phycological state under exercise. If you haven't read that post already I would suggest reading that one first since I won't cover all the same underpinnings again. However, whereas in zone2 one is interested in finding the first lactate inflection point-or LT1-in FTP determination and training, one is much more interested in the second lactate inflection point or LT2.

It is the point at which our workload has increased such that our system can no longer clear excess lactate from our working muscles, the mitochondria in our working muscles have no oxidative capacity to spare and the lactate concentration in our blood begin to run away out of control. In the literature this threshold is also designated by various other terms such as "MLSS (Maximal Lactate Steady State)", "OBLA (Onset of Blood Lactate Accumulation)" and "LT2 (Lactate Threshold 2)". Sometimes it is also called "anaerobic threshold", but this denomination is somewhat misleading.

Because our mitochondria become partially and acutely dysfunctional in these conditions, we become unable to push our power beyond this point for an extended period. This is therefore a "threshold" point, hence the use of "functional threshold power" in the "FTP" initialism.

The power that we can just barely sustain without exceeding this physiological bend in the road is our FTP and this is also sometimes referred to as our FTP TTE or Time To Exhaustion, which is literally the amount of time we can fight our way through this hurculean task before we become exhausted and either our power falls off or we stop cycling all together.

Why can't we sustain power just at ftp forever?

Fatigue is complex and multi-causal in origin, but what separates FTP from the other acronyms above is that it is a functional measure-that is, it is human-centred.

In contrast with the lab-based physiological measurements mentioned, functional power recognises that we are humans rather than machines, and that we may "give up" before we are strictly compelled to from a purely physiological standpoint. This is important, because it suggests gains in FTP can be reaped not just via physiological changes or improvements, but also in response to changes in environmental, situational, and psychological variables. The "functional" nature of FTP is highlighted by the fact cyclists will often find they have an "indoor FTP" and an "outdoor FTP", with different observed values for each, because of variant conditions. So how can I find my FTP power?

New riders may find it hard to get an accurate estimate of their FTP power, simply because they don't know how FTP "feels".

Consequently, they are not sure what power to target when attempting to ride for 40-70 minutes in a quasi-steady state. As a result, the so-called "ramp test" has become a popular alternative for FTP estimation, for beginners in particular. In this test, a rider holds a fixed initially low power for a fixed duration, e.g. 2 minutes. The rider then increases their power in set increments of power and duration defined by the test protocol, until reaching exhaustion; their FTP is calculated via a formula from the results of the test, typically by taking some percentage of their average final one-minute power. Since the power is dictated by the test protocol, no familiarity with FTP "feel" is needed-only a commitment to go till failure. Is the "ramp test" a good estimator of FTP?

Well, no. In fact, I would recommend its use only for very new riders who have no idea yet what their FTP is, or familiar with how riding at threshold feels.

Why does the ramp test not serve as a good measure of FTP?

To explain this, first it is useful to explain how a ramp test works.

We've touched on above that the ramp test relies on "riding to exhaustion", but this is actually quite an imprecise phrase. At least, it is possible to reach "exhaustion" in many very different ways, for example via a 15-second maximal sprint, or via a 24 hour ultra-endurance ride. You might even be fatigued from a 1RM deadlift, which may take less than 5 seconds. These are very intense forms of exhaustion or fatigue. The form of "exhaustion" or limit the ramp test is attempting to achieve is that at which the body reaches the maximum possible aerobic capacity,VO2Max, and also reaches some limiting factor in either the central or peripheral systems whereby it cannot ingest, absorb or utilize more oxygen than the cells require to produce energy aerobically.

Less scientifically: you will in all likelihood be gasping for air, feeling sick, and convinced that you might well die. The reason that a ramp test impresses this particular brand of unpleasantness upon us, is that from the resultant data it is possible to calculate our MAP or "Maximal aerobic power", often referred to as "power at VO2max".

So a ramp test doesn't actually measure my FTP!!!

Correct: a ramp test estimates your power at VO2max. It then attempts to "guess-timate" your FTP from this value. There are several reasons this is flawed.

First, the ramp test has not actually measured your power at VO2max and there is no way to know if you even hit your VO2max during the test - if fatigued or new to testing, you might simply have given up.

Second, for the short-duration test protocols, such as ramp tests, it is possible to "muscle through" the test using contributions from neuromuscular and/or anaerobic energy systems. This is especially true for new athletes who are typically anaerobically-dominant and hence may get an inflated MAP estimate.

The parameters of that test vary across different platforms-stepped length and step power increases. If such variables are not supported by available research, then the test may well prove highly erroneous. Particularly true of Zwift: it is a bit like saying "meter rules are an accurate measure of height", then producing your own meter rule which is 90cm long.

The figure calculated from the test is, fourthly - as we saw above, not your MAP power but your FTP.

It's related to a completely different physiological boundary, maximal oxygen consumption and not maximal lactate steady state-so it must therefore use "magic tricks" to attempt to deduce what your FTP is. It might do this through an often proprietary and hence un-validated formula that tries to do a "best fit" regression line against FTP from MAP derived on population-level data. It might say for example "FTP is 90% of your MAP power". The thing is, we are all different. Because there is no exact physiological relationship between MAP and FTP power, there is no reliable way this calculation can be made. Depending on which energy systems you have been training, this average percentage-that at the population level is OK-may lead to an estimate of FTP that is wildly wrong for you personally.

How do I go about testing my FTP?

As we saw above, the most comprehensive test protocol for determining FTP would be to determine your OBLA/MLSS/LT2 and observe power output at this level, all within laboratory conditions. In the field-such as in your own home-accurate field tests are possible using only a power meter-or turbo trainer-and a stopwatch.

If you can stomach it, attempting to ride at a power for around one hour in length-40-70 minutes, provides the best estimate of FTP. If you do this, then you may want to use a negative split for best results, riding at say 10 minutes a little under your target power, and then increasing effort gradually.

With this method you also take your average power for the test, not your normalised power - which can be higher. You should try to make sure as far as possible that you do the ride "steady state", i.e. evenly, not repeatedly spiking and dropping power.

I can't face riding for an hour flat out

I would say "are you sure"?.

Have you tried? It's a great workout in and of itself, and there is even a phrase in cycling: "embrace the suck". It will give you the most correct measure of your FTP, and it is going to yield benefits in terms of mental toughness that is so important in cycling. And in addition to all that, it's going to be the best indication of how riding at threshold "feels", which is critical for dialling in your perception of effort. Try it-you may surprise yourself! But if you really can't face it, then a next-best option may be to undertake a 20-minute test, but try and do it as per the originally designed protocol: do 5 minutes flat-out at VO2max, rest for 5 minutes, then do 20 minutes flat out at a steady power output as you can manage, take the average power from that 20 minutes, and take 95% of that 20 minute power as your FTP.

Just keep in mind you are, as in the ramp test, "guess-timating" your FTP still, albeit to a considerably lesser extent. Why is it important for me to know my FTP accurately?

It may not be.

Or rather, it may not be any more significant to you compared to any other point on your PDC. "My FTP is N" does have the advantage in that it gives a quick and easy indication of your overall performance, useful in conversation with others. Unlike zone2 training, threshold training has no specific benefits to overall health, and it also isn't necessarily the most critical power/duration for you to train even from a performance standpoint, even if you race. As we discussed from a physiological perspective, FTP power is related to the LT2 inflection point; however, this particular point on the PDC may simply not be too critical to you if it does not have specificity to your required performance outcomes.

If you race and those races are highly stochastic-for example-then aerobic power or repeatability of effort may be more important. This would also be true if you target ultra-endurance riding, in that your LT1 or zone 2 power may be more critical in determining your success. It is useful in training to have some idea of your "SWOT" strengths, weaknesses, opportunities and threats, and to base your training around these. The importance of FTP is, therefore, potentially over-emphasised for many riders.

How can I know I'm riding at my FTP power?

In the lab we could try to estimate your OBLA/MLSS/LT2 as a physiological parameter and measure power output at that point. In the field we don't have access to this information. Our functional power on any given day can be estimated in a variety of ways, including a guess based on our last FTP test and our recent training and recovery history since that point.

But most accurate of all is simply "feeling" that you are riding at threshold. That is why I advise riders get used to riding flat out for an hour at least a couple of times per year-so that they know what maintaining threshold right at that burn point feels like, and can ride at this power when needed in racing.

There are formulas to estimate if you're in your threshold power zone based on heart rate-either based on historical heart rate seen during threshold riding, or as a percentage of your max heart rate-but I'd avoid these since HR is highly variable on a daily basis based on various factors including heat and fatigue.

What is a good FTP?

Attempts have been made to estimate what "typical" and "good" FTPs are at a population level, and Zwift uses FTP for rider categorization, but overall you shouldn't worry too much if your FTP is "good" at a population level-only what it was in the past and how you can improve it to meet the demands of your particular objectives. Work on your FTP power specifically if this is needed for your targeted goals.

So, should I use my FTP to set all my training zones?

No. Many training platforms - Zwift included - are guilty of this. As we saw in our earlier post on zone2 we use a 7-zone model in cycling training for both description and prescription of session or interval effort. These zones align with different energy systems and physiological parameters and few of them are related to your LT2 inflection point.

Just like when we discussed the ramp test, taking a fixed percentage value of your MAP/VO2Max power - zone5 - and declaring that is the value of your FTP power - zone4 - is not safe. And vice versa, you cannot correctly take a percentage of your FTP - zone4 - and declare this to be your VO2max - zone5 - or endurance power - zone2. This, of course, is because we all have unique PDCs.

There exist some test protocols aimed at better defining rider-specific power profiles: "WKO" software will attempt to model your PDC based on natural in-ride efforts-so long as you'll delivery maximal efforts at difference durations in order for it to be accurate; "Critical Power testing" is an alternative testing method to FTP testing which attempts to capture power generated at different durations via different energy systems; and Wahoo's "4DP" test is a structured test protocol aimed at capturing neuromuscular-depositive (5 second), anaerobic (1 minute), MAP (5 minute) and threshold (20 minute) power points.

Zwift is behind the curve on this currently. How accurate are training zones based on a % of FTP (as in Zwift)?

Not very, especially for zones above FTP (zones 5,6,7).

To show why, lets consider sprint power (zone7) for two riders. Both riders have an FTP of 300w. One rider has an 800w sprint, the other a 1400w sprint. But if we try to base zone7 (sprint) power for prescription in training on a fixed percentage of FTP, how can we possibly do this when riders have divergent power duration curves? This holds also true for other training like VO2max/MAP-zone5 and "anaerobic" training-zone6. Below threshold, there is somewhat more commonality between athletes, but even so we'd recommend testing zones/energy systems independently and basing them as much as possible on some of the underlying physiology and energy systems that they relate to in order to optimize your training .

Should I train at threshold to increase my FTP?

As we have discussed in our blog post regarding zone2 training, one doesn't necessarily have to ride in a given zone to improve one's power in that zone. So too with FTP - one does not need to ride at the FTP power in order to improve one's FTP power.

This might sound a little backward, but bear with me. To understand how this all works in the first place, we have to go back a step and recognize what "training" even is. We train for physiological adaptation. In other words, we put some sort of stress on the body to which it is unadapted. In response, the body attempts to make some adaptations to our physiology in order to better prepare us should we encounter similar situations sometime in the future.

By training, our body adapts over the days, weeks, months, and years by increasing blood plasma volume, our glycogen storage capacity, muscle capillarization, muscle fibre type balance, and cardiac stroke volume. For example, the hearts of professional cyclists are much bigger and their blood volume much greater after many years of training.

However, adaptations which are triggered by the body in response to training in one zone, may be reaped as benefits to power uplifts in other zones: it is not necessary to do all-or even very much-of our training at threshold power in order to increase our threshold power.

Contrary to that, even when our primary goal is to enhance our FTP level, most of our rides, approximately 80%, could be more effectively done in tempo zone3, endurance zone2. Therefore, the question is, for how long should I train my FTP?

If we wish to improve our FTP/threshold power, for whatever reason, we can prioritize optimizing our FTP power for a 4-8 week period in a harder block of training, leading into a key target event we would like to peak for. This could be part of broader macro-periodisation under an annual training plan (ATP).

How soon will I realize gains?

Gains to FTP are realized reasonably quickly to begin with, as the benefits derived from some of the adaptations (particularly neuromuscular adaptation and some cardiovascular adaptations) come relatively quickly - especially if you have no or limited prior training history. Other physiological benefits continue to be gained at a much lower rate, over several months and even years of concerted training.

How many Threshold workouts per week should I do

We want effective training to depend on a period of overload-whereby we put a new stress on the body-followed by a period of super-compensation or adaptation, whereby we reap the rewards of that training as our body adjusts to better prepare for similar efforts in the future.

The structure of these hard training-overload and recovery-or reduced training-super-compensation and adaptation-phases exist at various scales from day-to-day, week-to-week, and month-to-month. The structure we pick is termed periodisation. There are many approaches to this-we shall cover two here, both depending on the approach taken to periodisation. Classically, periodization is the process by which we systematically change the prescription of workouts in some sort of repeating cycle of uptick via steps, typically weeks, where we increase the load of training each week for 3-4 weeks, going harder and harder each week before backing off and recovering, repeating the process.

We might do two threshold workouts per week, separated by days of easy riding or days off the bike. But in each workout we can come in with the idea to push up by hard-pressing for more power output, or push out for longer duration, or both. A variant is block periodisation where we take all of the "intensity" workouts from a planned 3-4 week classic periodisation cycle, and jam them all together into one week, then ride easy for the remaining weeks in the block. The theory here being that because there is a heightened overload phase, the increased intensity of signalling yields greater gains. If this sounds tough, its because it is.

How do I structure my threshold workouts?

Threshold workouts are all performed at or around the threshold power.

The exact workout prescription depends on whether we are most interested in improving our threshold power aka push "up" or increase our TTE/time to exhaustion at threshold power aka push "out". If we can hold our threshold power for 40 minutes only in testing, then we likely want to work on TTE. Meaning if we can maintain our FTP power for 60-70 minutes we probably want to move it "up". This is another reason, as mentioned earlier I feel full-hour testing of the FTP is useful: it shows which one of these is the biggest limiter in a given athlete at a given point in time. Once we have isolated our need-increased power or duration-we can construct a progression of workouts that gradually increase the needed metric in a series of workouts over the course of our training block.

Since riding at or around FTP is fatiguing, it is most common to perform threshold training using intervals-or periods of work and rest-within a given session.

What kind of interval structure should I employ?

The two most typical threshold interval structures are:

Steady state intervals, where we look to do repetitions of work intervals at or around our FTP.

Structured as for example 4x10 minutes at 100% FTP power. That is, 10 minutes at 100% FTP, then 5 minutes rest, and then repeating both work and rest in that sequence 4 times, for a total workout length of around one hour. These are simple and effective, provided the principle or progressive overload is followed-more about this below. The second common form is over-under intervals in which we alternate riding over and slightly under our threshold.

There are lots of variants, but they generally tend to be based on a desire to prioritize the generation and subsequent clearance of lactate, and that this lactate clearance is trainable. Such intervals would typically involve a short period- say, 30 seconds-in zone 5VO2max power-followed by a longer period-say, 2 minutes-in zone 3. A series of 8 of these oscillations without rest counts as a single "block" equivalent to 10 minutes of steady riding in zone4. We might then rest for say 5 minutes, before performing the oscillations and rest in sequence another 3 times. When doing "over-unders" specifically to target lactate clearance there is some evidence optimal lactate clearance may occur in mid-tempo range, which in some athletes may occur around 80% of FTP power.

How long should my intervals be?

The duration of a threshold interval should typically be no less than 10 minutes for a threshold interval at 100% of FTP, with a typical maximum of around 30 minutes.

If we were working on increasing our TTE, we might increase either the individual length of each interval in each subsequent workout (4x10 -> 4x11 -> 4x12) or else add in additional intervals (2x10 -> 3x10 -> 4x10), or some hybrid of these. The important metric when training TTE is total Time In Zone (TiZ).

If we are working on increasing our power at FTP then we might, by contrast, stick with the same duration (4x10 mins) through our training block but increase the percentage of FTP power that we hold for each block in each subsequent workout session in a block, e.g. 100% -> 101% -> 102%.

What power should threshold intervals be at?

The length of the threshold interval is inversely proportional to the power output.

It's entirely possible to do threshold intervals either at a higher or lower power output, providing we stay inside the bounds of zone4 (ish) and adjust power output accordingly. We might for instance replace our 4x10mins at 100% with "3x20mins at 95%", or "4x8mins at 103%". This is also effective as long as we do stay in zone long enough to deliver effective stimulus and stay predominantly at a power around threshold. Should I use ERG mode for threshold workouts?

I don't think that's a problem to do threshold workouts in ERG mode and perhaps even beneficial, just not for testing FTP. Research-wise, I'm not aware of any that would show one way is better than the other, so just so long as you hit your targets, feel free to do either.

Can I decrease the power if I'm tired?

It will of course remain an effective training dose of the workout if you keep in the prescribed zone and reduce power up to 5% within reason. However, if you are in a block of progressive training, you will of course have "failed" with progression. Normally this is an indication of over-fatigued status.

Myth: What happens if I "fail" one threshold workout?

If you fail a threshold progression, my first suggestion is simply to assume that this is merely a "bad day"; take a couple of days off the bike, rest up, refuel and hydrate. Then, try again with the workout. If you smash it the second time, keep going along in your training plan. If you fail again, then it is worth considering that you may be overfatigued more chronically and need a longer period of rest.

How do I optimize individual training sessions of FTP?

Ensure you start the training session adequately fuelled - that is your glycogen stores are not depleted. This means making sure to eat plenty of carbohydrates in the 24 hours previous to your threshold workout, especially if having trained hard the previous day.

As we saw earlier, it is beneficial to "prime" our aerobic system's slow component before a threshold effort.

There is some disagreement as to how effective this is, but you may want to try preceding threshold training intervals with a few short periods of high intensity - say one minute at FTP power, or 30 seconds at VO2max, repeated with rest. For some individuals, this has the added value of reminding them mentally how the threshold effort will "feel".

How would I optimise my FTP training in a macro annual sense?

It is not possible to simply increase threshold power via threshold workouts indefinitely. In particular, threshold power can never exceed our max aerobic power VO2Max. Once you stop seeing significant threshold gains you may therefore wish to refocus on VO2max increases for a period, via zone2 chronic adaptation or zone5 acute adaptation training blocks.

I find threshold intervals hard; do you have any tips?

Threshold training is mentally tough and involves some element of protracted suffering. Performance may be determined as much by psychological state as physiological state. Various techniques have been proposed for improving our ability to push through tough threshold workouts, including "chunking", in which you break down a workout that might seem daunting into smaller, manageable chunks.

Second, psychologists outline various types of attention focus.

It is worth experimenting with what works for you during a workout-trying to "escape" the discomfort of a workout by choosing an external focus, distracting ourselves from the effort, or using a "narrow internal" focus, deliberately focusing on the burn and how it feels. This latter relates to the concept of mindfulness: that we do not have to try to eliminate the discomfort but simply acknowledge it without being perturbed by it-learning to recognize that it is not harmful but part of the positive process of training. Third, many people find that they can go harder if they say some kind of mantra in their head. For some, this is: "I'm doing this for my health", for others: "pain is only temporary". Just find a mantra which is personal and motivating for you.

Fourth, studies have proven that music can actually lessen our subjective perception of effort. Some individuals save tracks to motivate themselves more for the FTP tests or threshold workouts.

Fifthly, you may want to try and find out common factors of success for the better workouts you have done: do you normally do better workouts at the beginning or end of the day?

Sixth, though there are extremely few proven-legal-ergogenic aids, one does stand out above all others: caffeine. Dosing is important, and effectiveness in an individual does-like so much else-depend on our genes, but it may be worth experimenting with ingesting caffeine around 15 minutes before your workout begins.

Seventh, fuel your threshold work with carbohydrates and simple sugars in particular: the nature of threshold riding is highly glycogen-dependent. Fuelling during a workout can increase performance even if we are fully fuelled before starting our ride: studies show that the simple "mouth feel" of sugary drinks can boost performance even if the drink is spat out.

Lastly, if indoors, ensure you have sufficient active cooling in the form of fans and aircon. Overheating can hold you back as your body dials back power output.