Shindo What zone 2 is, and what it is not#
Zone 2 is low-intensity aerobic work: sustainable for hours, conversational, sitting below the first ventilatory threshold — the intensity at which breathing first becomes noticeably deeper and more structured. In lactate terms it is usually placed around 2 mmol/L, which is where the widely-used name comes from.
Two clarifications, because both are routinely lost.
It is a physiological boundary, not a slice of your maximum heart rate. The five-zone charts printed on treadmills define zone 2 as, say, 60–70% of maximum. That is a different definition that happens to share a name, and on many athletes the two land in genuinely different places.
It is defined by the ceiling, not the floor. The training effect comes from time spent below the boundary. Going slower than necessary costs you a little stimulus per hour; going above it converts an easy session into a moderate one, which is the actual failure mode — see why the middle is the problem.
Why the usual formula misses#
The default method is to estimate maximum heart rate as 220 minus your age and take a percentage of it. Both halves of that are weak.
The age formula is a population regression with an error band of roughly ±10–12 beats, which for zone work is wide enough to place an individual in the wrong zone entirely. And the percentage that corresponds to the first threshold is not fixed: in trained endurance athletes the boundary sits proportionally higher than in untrained ones, and it moves as you train. Two people with identical maximum heart rates can have first thresholds twenty beats apart.
The practical consequence is specific and common: an aerobically well-developed athlete following the formula trains their easy days too slowly and their moderate days in a zone that does not exist. There is more on this in heart-rate zones from your own data.
Three ways to find it from your own data#
None of these is a laboratory measurement. Each of them uses information you already have, and each has a known weakness.
1. Breathing and speech#
The oldest method and still the most robust, because it observes the thing the zone is defined by. At the top of zone 2 you can speak in full sentences but would rather not give a speech; your breathing has a rhythm you notice but do not have to manage. The first threshold is approximately where sentences start breaking into fragments.
Weakness: it is subjective, it drifts when you are tired, and it is unusable in a group where conversation sets the pace instead of reporting it.
2. Anchor to a modelled threshold, then take a fraction#
If you have a spread of maximal efforts in your history — a 5-minute effort, a 20-minute one, a longer time trial — you can fit a power-duration or speed-duration curve and read off critical power or critical speed. That is the second threshold, not the first, but the two are related closely enough to be useful: the top of zone 2 usually lands somewhere around 75–80% of critical power, or roughly 80–85% of critical speed in pace terms.
Weakness: it is an inference from another inference. The fraction varies between athletes, and the curve is only as good as your hardest efforts — if you never go deep, the model has nothing to fit.
3. The drift test#
The strongest field evidence, and the one most people have never run. Do a long steady effort — 60 to 90 minutes — at a candidate intensity, then compare the first half with the second half and calculate decoupling: how far heart rate drifted relative to pace or power.
Under about 5% means the intensity was genuinely aerobically sustainable. Substantially more means you were above the boundary, whatever the heart-rate chart said. Repeat a couple of weeks apart, adjust the candidate intensity by a few beats or watts each time, and you converge on a boundary derived from your own physiology behaving over an hour rather than from a formula.
Weakness: it needs flat, uninterrupted terrain and stable conditions. Heat alone can produce several percent of drift and invalidate the reading — which is precisely why the result is worth having, since the same heat is distorting every other estimate you make. (Heat and altitude bend pace in ways worth understanding on their own — here is the correction.)
What “without a lab” honestly costs#
The title of this guide is not a claim that field methods match laboratory testing. They do not, and the differences are worth stating plainly.
A lab gives you a directly measured quantity — blood lactate, or expired gas — rather than an inference from heart rate and pace. It controls temperature, hydration and protocol. It gives you a defensible number on the day.
What it does not give you is durability across the season, a reading taken while you are tired, or any information about the other 364 days. A boundary estimated from months of your own training is noisier per observation and considerably better sampled. For deciding how fast to run on Tuesday, the second is usually the more useful object.
Treat any field estimate as a band of several beats or watts, not a line. Then train below the bottom of the band and stop worrying about it.
When your history is too thin#
All three methods need material. If you have four weeks of data, one long ride and no maximal efforts, an honest system has nothing to fit — and the correct behaviour is to say so rather than to produce a confident-looking number from an age formula.
What to do in the meantime:
- Use breathing. It works from day one and needs no history at all.
- Deliberately create the inputs. A 20-minute time trial and one long steady effort, a fortnight apart, give a threshold model something to work with.
- Assume the boundary is lower than you would like. People overestimate their easy pace far more often than they underestimate it, and the cost of being 5% too slow is much smaller than the cost of being 5% too fast.
- Distrust the first month of any zone estimate, from any source, including this one.
The mistake to avoid#
Treating the boundary as a hard line and the session as a compliance exercise. Zone 2 is a region, its edges are soft, and the number was uncertain to begin with. Ten seconds a kilometre either side of your estimate does not change what the session does to you.
What does change it is a hill, a headwind and a habit of pushing back up to the target number on every rise. If you are staring at a watch and correcting continuously, you are running an interval session with the intensity of an easy one — the worst combination available. Set the ceiling, run under it, look at the number afterwards.
Shindo derives thresholds and zones from your own training history rather than from an age formula, models critical power and W′ for cycling and critical speed and D′ for running and swimming from the best efforts in your data, and reports decoupling on long steady efforts. It reads sessions your watch has already written into Apple Health — it does not record them, and it measures no lactate. Threshold and zone modelling sits in the paid tier; readiness, training load and your weekly plan, over your most recent 60 days of history, do not. The methodology sets out what is modelled.
Read next