Cycling training zones: how to set and use them.

How do you set cycling training zones?.

• Start with the system your plan uses. A target based on FTP is not automatically the same as one based on CP, maximum heart rate or a measured threshold.
• Check the anchor. Use recent, well-executed test data. Our Critical Power testing guide explains the inputs for the calculator.
• Check the response. Watts describe external work; heart rate and perceived effort help you interpret the cost. Heat, fatigue and the duration of the effort affect that response.
• Adjust with a purpose. An endurance ride should remain appropriate to its intended effort. A target on the screen does not prove which physiological domain you are in.

What does Zone 2 mean?.

In many cycling plans, Zone 2 describes endurance riding. The exact boundaries depend on the model. Comfortable conversation and a manageable effort are practical cues, but they do not measure a lactate or ventilatory threshold. CP alone does not locate your first threshold either.

Below, I explain my four-zone coaching framework and the research behind treating percentages as starting estimates. For the difference between the power anchors, read Critical Power vs FTP.

The zones problem.

Right. Training zones. And this is where I'm going to annoy some people. I know it.

If you've been cycling for any length of time, you've got zones. You know what zones are. You know what they're supposed to do. And you may be just as religious about zones, if not more, than power. You probably either use 5 of them, maybe 7. They're set as a percentage of your FTP. Zone 1, you think about recovery. Zone 2, you think about endurance. Zone 4, threshold. And somewhere in there you've got a sweet spot that everyone argues about on the internet.

Here's the problem. Those zones, the ones you're probably using right now, are based on fixed percentages of a single number. And the research is pretty clear that this doesn't work the way we think it does.

Jamnick and colleagues' 2020 review examined how well different methods of prescribing intensity correspond to physiological exercise domains. They questioned fixed percentages of maximal anchors such as maximal oxygen uptake, maximum heart rate and maximum work rate.

For coaching, the useful question is whether the target produces the response intended for this rider.

Why fixed percentages fail.

The fundamental issue is that two riders with the same FTP will have different lactate thresholds, different fat oxidation rates, different anaerobic capacities. Setting their zones as the same percentage of FTP ignores all of that individual variation. You're getting approximate zones at best and meaningless ones at worst.

Now the Coggan model is everywhere. It's built into TrainingPeaks, it's the default in most platforms. Seven zones, all derived from FTP. And it was useful when we didn't have better tools. But we do now. And I think most people are using yesterday's system because it's the one that comes pre-loaded.

But here's the thing that really matters, and this is where it gets interesting. Underneath all the zone models, Coggan's, Seiler's, any of them, there are real physiological boundaries called exercise intensity domains. Moderate, heavy, severe. These are genuine thresholds where fundamentally different things happen inside your body. Below CP, your oxygen consumption and lactate can stabilise, you can reach a steady state. Above CP, they can't, they rise progressively until you're done. That's real. That's physiology, not a coaching model.

But here's the problem: those boundaries move. Stevenson and colleagues showed this in 2022. They tested trained cyclists at rest and found their first ventilatory threshold, the boundary between moderate and heavy exercise, which is roughly your Zone 2 ceiling. Then they had them ride for two hours at 90% of that threshold, and tested again. On average, the power at that threshold dropped by about 10%. Heart rate at the same threshold rose by about 6%. And the individual variation was massive, drops ranged from roughly 4% all the way to 20% across the group.

What that means in practice is this: if you're riding at the top of your Zone 2 based on a test you did fresh on a Monday morning, by two hours into that ride, you might actually be in Zone 3. The boundary has shifted underneath you. Your zones have literally changed during the ride, and the software you're using doesn't know that.

So this is why I still use percentages when I prescribe. Because I need to give you a number before you clip in. But they're a guide, not a border. I think of zones as neighbourhoods, not postcodes. The physiology is a spectrum, and it's shifting all the time. Even CP itself changes within a single ride.

And the effective way to deal with that isn't to constantly recalculate your zones every hour. That's madness. It's to adjust based on the response. Read what's happening, power, heart rate, RPE, and respond to it in the moment. That's why triangulation matters. That's why we keep coming back to those three signals together. The zones get you in the right neighbourhood. The response tells you whether you're actually where you need to be.

TrainingPeaks and most software use fixed zones for the whole ride when calculating time in zones. So your post-ride analysis might show you spent two hours in Zone 2 when you actually crossed into Zone 3 an hour in. The report doesn't match the physiology.

What the interview adds: read the effort, not just the zone.

In our Cycling Performance Club conversation, exercise-intensity researcher Nicholas Jamnick explained that exercise domains describe physiological responses. A zone label on a screen does not establish which response a rider is experiencing.

His 2020 review with colleagues evaluates intensity prescriptions against moderate, heavy and severe exercise domains. It supports critical power as a way to distinguish the heavy and severe domains, while also noting that immediate exercise responses do not establish long-term training adaptations.

That gives us a useful question after a ride: what happened during the actual work?

Same average power, different session.

Here is a simplified example, created for this explanation:

• Ride A: 30 minutes continuously at 200 watts.
• Ride B: 20 minutes at 150 watts, then 10 minutes at 300 watts.

Both average 200 watts. For Ride B, that is (20 × 150 + 10 × 300) ÷ 30. The average has hidden the change in intensity. It cannot tell you that either ride stayed in one exercise domain, and we cannot assign domains from these wattages without knowing the rider.

When you review intervals, look at the work periods and recoveries separately, alongside heart rate and perceived effort. Keep the session average as a summary, but use the detail to assess how the work matched its purpose.

If you are starting with a number from the Critical Power calculator, read the testing guide to understand the inputs. The CP versus FTP chapter explains the distinction between those anchors. Calculated zones remain starting estimates; CP alone does not locate every physiological boundary.

What I use instead.

So I went through a process with this. I looked at every zone model available. Coggan's seven, Seiler's three, Skiba's CP-based levels, Issurin's block periodisation zones, the ISM model, the Italian model. I've got a spreadsheet where I mapped every single one of them against each other. Every metric, watts, watts per kilo, percentage of max, percentage of CP, heart rate, blood lactate, RPE, duration, fibre type recruitment, endurance methods. Everything.

And I'm showing you this because I want you to see what the process looks like. This is the work that sits behind every decision I make for an athlete. Every coach should have done this, sat down, mapped the options, compared them against the physiology, and made a deliberate choice. Not just used the default because it came pre-loaded in the software. Ask your coach if they've done it. Not to be a prick about it, but because you deserve to know whether your training is based on something, or whether someone just picked the model that was already there.

What became clear is that most of these models are describing the same physiology with different labels and slightly different boundaries. The arguments between them are mostly academic. For a coach working with a real person, the question isn't "which model is correct?" It's "which model is useful?"

And I landed on something simpler than any of them.

I use four functional zones. Not because the body only has four intensity levels, it obviously doesn't. But because for training prescription and execution, four is the minimum number that changes your behaviour without creating false precision.

Zone 1: Recovery. I don't rule this out as being endurance light, there is some benefit for some time rolling around in this, it's still adding some stress to the aerobic system. But the reason I keep the label as recovery is a coaching cue. I want to push people down and keep them there. If your heart rate is drifting up or you're having to think about the effort, it's not recovery. The label does the work.

Zone 2: Endurance. Your aerobic development zone. This is where the bulk of your training lives. Bounded by your actual aerobic threshold, not a percentage of FTP. Set it where your heart rate stays stable and you can hold a conversation. If those cues change, reassess the effort; they do not establish a physiological threshold on their own.

Zone 3: Tempo. And I think it's important to have a clear delineation here, this is tempo work, not threshold. Tempo has its own purpose and its own feel. It's structured, sustained effort below your threshold. Lumping it in with threshold blurs the intent.

Zone 4: Threshold and above. This is everything from CP downward, the lower end of threshold, right through to above CP. And here's the key choice I've made: everything above threshold is open. It varies by duration, by the type of effort, whether that's aerobic sprinting, the so-called VO2max intervals, whatever you want to call them. We need a better name for VO2max intervals, honestly, because it's on the athlete to find that intensity, not a percentage.

And here's an example of how I'd coach someone through those shorter intervals. Say it's 6x3 minutes. Set 1, hit the target. Sets 2 through 5, go on feel. Set 6, finish strong and finish empty. No percentage of any zone can replace that type of effort. The coaching is in the cue, not the number.

I'm going to drop coaching nuggets like this through the playbook as we go. It's more tactics than strategy, but there's value in the detail.

Why I chose this.

I want to explain why I've deliberately gone simpler when most of the industry is going more complex. There are two main reasons.

First, precision is only useful if it changes behaviour. If I give you seven zones and you can't reliably tell the difference between Zone 3 and Zone 4, or Zone 6 and Zone 7, when you're riding, those extra zones aren't helping you. They're just creating anxiety around the workout. Four zones means you always know where you are. You're easy, you're building, you're working, or you are flat out. And that is enough.

Second, and this is the one I feel strongly about, each zone should be derived from its own relevant metric. That's not meant to sound complicated. In other words, not percentages from one number. Your endurance zone should come from your actual aerobic threshold. Your threshold zone should come from your actual Critical Power. Your VO2max area should relate to your actual VO2max or your actual power at that intensity, not 106 to 120% of something else.

This is where the Mastersheet comes from. I've built a zone system where every zone is anchored to the physiology it's meant to target. It takes more work to set up than just plugging in your FTP and letting the software do percentages. But it means your zones are actually yours, not a statistical average that might or might not describe your body.

Now, separate to this, and I might get called out on this, I still prescribe within TrainingPeaks with percentages. The big thing that I have to get through is that there's a lot of responsibility on the side of the athlete to make sure that they understand themselves and that they know when to adjust. I spend a lot of time on this, and I'm sure we'll get into it later. I'm not doing an inside test, any type of metabolic cart test. I don't actually have a way to find LT1 or the aerobic threshold, so I rely on the anchor from CP, and then it's up to the athlete. I make prescriptions for different levels of intensity within Zone 2 based on what I'm trying to achieve, so we can move people around the zone a little bit, not just give them this huge wide range, and the sensations under that should feel different.

How to set your zones properly.

You need data, not just a test. A single 20-minute test gives you one number. To set zones properly, you need your Critical Power from multiple efforts.

For heart rate zones, I don't go out to specifically set heart rate zones. I keep an eye on if a max heart rate appears in the data, and when it does, we double down on that and set from there. The research shows that heart rate at the first threshold can sit anywhere from 60% to 90% of max depending on the individual. Iannetta and colleagues showed that across 100 subjects in 2020, so percentage-based HR zones are always an approximation. But max heart rate is the most practical anchor I've found, and it's where my approach originally comes from.

Beyond that, it's on the athlete in the session to understand what happens at different intensities. If you're decoupling on a ride, heart rate drifting up at the same power, that tells you something. We get into that. But I'm not sending someone out to do a max heart rate test or a two-hour stability test. I look at the overall trend of what the heart is doing, not whether it's sitting neatly in some zone.

Connecting zones to fuelling. Each zone has a fuelling demand. Recovery and endurance, you can fuel primarily with fat plus moderate carbs. Tempo and above, you need carbs and the demand scales with intensity. Fuelling prescriptions per zone aren't a general rule. There are many different ways you can approach fuelling and it's very individual. Some riders get low, some riders get high, some riders get in between. It really does vary. But it is important to note: if you've never seen a chart where you see the fuel usage underneath the zones, you're in for a big shock. The amount of glycogen fuelling threshold compared to moderate Zone 2, it's that much of a shock that it will make you rethink your nutrition, most likely.

Zones change over time. Your zones may change through a training block, but unchanged threshold power does not prove that training has failed. Reassess regularly. Not obsessively, but every six to eight weeks at minimum.

But I'd say that with a caveat. This is the danger of just one number for an entire system. Some things can change, especially when you're leaning more durability-focused, without your threshold number changing. So you need a way to pick up other things as well.

And look, the broader point here isn't that wrong zones will mess up some underlying physiological process. These are on a spectrum, like I said at the beginning. It doesn't necessarily matter for the physiology underneath. It's the practical stuff on top where you start feeling the pain. Riding harder than you need to. Giving yourself false confidence when you're feeling strong but really it should be the other way around. Extra fatigue because the fuelling is off, because the intensity was off, because the zones were off. There's a whole chain of consequences, and they're all practical, not theoretical.

Zones are a tool, not a cage.

Last thing on this, and it might be the most important thing I say in this entire chapter.

Zones are a framework. They're a language for describing intensity. They're not the ride. The ride is what happens when you clip in and go. Zones help you make better decisions about how hard to go and when, but they don't replace listening to your body, reading the signals, and making a call based on what's actually happening.

I've seen riders turn themselves inside out trying to stay in Zone 2 on a day when their body was telling them to rest. I've seen riders refuse to push because "it's a Zone 2 day" when everything was lighting up and they had a great session in them. The zones serve you. You don't serve the zones.

That distinction, between using a system and being controlled by it, that runs through everything I'm about to teach you. And it starts here.

Right. Now that we've got the language sorted out, power, your curve, your zones, let's talk about the thing that ties all of this together. The layer that almost nobody is teaching. Durability.

When you need help with your zones.

If your targets and the way the sessions feel keep disagreeing, the next step is to review the training in context. See SEMIPRO’s cycling coaching options for help interpreting your results and deciding what to change.