Low-intensity aerobic training has been a staple of endurance coaching for decades. It reached general fitness discussion recently under the label zone 2, and in the process acquired both enthusiasm and confusion.
What the zones are
Training zone models divide intensity into bands, typically defined by heart rate, power output or blood lactate.
Different models exist with three, five or seven zones, and they don't map onto each other cleanly. When somebody says zone 2 without specifying a model, the term is ambiguous.
In the usage that's become common, zone 2 refers to a moderate intensity below the first lactate threshold — where blood lactate remains near baseline and the effort can be sustained for a long time.
Practically, it's the intensity at which you can hold a conversation in full sentences but would prefer not to. Breathing is elevated and controlled. It generally feels easier than people expect, and that's the most common error — training too hard.
The physiological rationale
The claimed benefits centre on adaptations that occur preferentially at lower intensities.
Increased mitochondrial density and capillarisation in the muscle. Improved capacity to oxidise fat as fuel. Improved cardiac output through increased stroke volume.
These adaptations are real and well established in the exercise physiology literature. The claim that they occur preferentially at low intensity is more debated — high-intensity training also produces mitochondrial adaptations, sometimes more rapidly per unit of time.
The stronger argument for low-intensity work is about volume rather than superiority. Low-intensity training can be sustained for far more total hours per week without accumulating fatigue, and total volume appears to be a primary driver of endurance adaptation.
The polarised training evidence
The underlying research most often cited concerns intensity distribution in trained endurance athletes.
Observational analyses of elite endurance athletes have repeatedly found that a large majority of their training volume is performed at low intensity, with a smaller proportion at high intensity and relatively little in the middle.
Studies comparing distribution models have generally found polarised approaches performing at least as well as threshold-focused ones in trained athletes.
Two caveats matter for general readers. This literature concerns trained endurance athletes training many hours weekly, and the distribution question is about how to allocate a large volume. Somebody training three hours a week is in a different situation entirely, and for them the evidence supports high-intensity work having a favourable time-efficiency profile.
Finding the intensity
Several methods, in descending order of accuracy and ascending order of practicality.
Lactate testing. The reference method, requiring blood samples during a graded test. Accurate and mostly done in laboratory or clinical settings.
Ventilatory thresholds from a gas exchange test. Similarly accurate and similarly requires a facility.
Heart rate percentages. Widely used and quite imprecise, because population formulas for maximum heart rate have substantial individual variation. Using a formula-derived maximum introduces error before you start.
The talk test. Unfashionable, free, and reasonably validated against physiological thresholds in several studies. If you can speak in complete sentences but would rather be quiet, you're approximately in the right place.
Nasal breathing. Some practitioners suggest breathing through the nose as a constraint that naturally limits intensity. Not validated, and it does impose a ceiling.
For most people the talk test is adequate. The precision gained from anything more elaborate is smaller than the variability introduced by sleep, heat, hydration and stress on any given day.
The main practical error
Going too hard, and it's nearly universal.
Intended easy sessions drift into moderate intensity, which is hard enough to accumulate fatigue and not hard enough to drive high-intensity adaptations. The result is chronic moderate fatigue and limited progress in either direction.
The correction is uncomfortable, particularly for anyone who equates training with effort. Genuinely easy sessions feel unproductive, and running or riding slower than you're capable of is psychologically difficult in a way that going hard isn't.
Practical tip that helps: train easy sessions alone or with someone slower, and stop looking at pace. Pace-driven training makes easy days hard, reliably.
Realistic expectations
For general health, the evidence for moderate-intensity aerobic activity is well established and doesn't require any zone framework. Official guidelines are expressed in minutes per week at moderate or vigorous intensity, and meeting them is the substantive thing.
For endurance performance, distributing training with most volume easy and a small proportion hard is well supported, particularly as weekly volume increases.
For someone with limited time, higher-intensity work is more time-efficient for cardiorespiratory fitness, and the trade is that it's less pleasant and carries more fatigue.
The framework is useful. What it isn't is a novel discovery, and the enthusiasm has occasionally outrun what the underlying research claims.
General information only. Consult a healthcare professional before beginning a new exercise programme, particularly if you have a cardiovascular condition.
Recovery is the other half
An element frequently lost in discussion of training distribution: the reason elite endurance athletes can accumulate enormous low-intensity volume is that their lives are organised around recovery.
Sleep, nutrition and the absence of a demanding job are part of what makes that volume tolerable. Somebody applying the same distribution around full-time work and family responsibilities is operating under different constraints.
Which argues for adjusting expectations rather than the principle. The distribution logic holds; the volume does not transfer. A person training five hours a week should probably prioritise consistency and adequate recovery over optimising the ratio between intensities, because the binding constraint is different.