How much of your running should be easy?
Most of it. At least 70% of training was easy in every successful approach that has been studied, and roughly 75–85% is a sensible target. Faster marathoners don't do more hard running than slower ones. They do far more easy running.
What to do
- Aim for roughly 75–85% of your running time to be easy. Treat that band as a working target, not a measured optimum.
- When you have more time to train, add it as easy running. Keep the hard work about the same.
- Run easy runs genuinely easy: you should be able to hold a full conversation.
- Count your split by time, not by number of sessions, and stick to one way of counting so you can compare weeks.
- Avoid a week where most runs are moderately hard. Threshold-heavy training did worse than mostly easy training.
What “easy” means here
Most research splits running into three zones, set by two thresholds. A threshold is an intensity where your body’s response changes. The first one is where lactate, a by-product of burning carbohydrate, starts to rise above resting levels. The second is where lactate starts to build up faster than you can clear it.
- Zone 1 (easy): below the first threshold. You can talk in full sentences.
- Zone 2 (steady): between the thresholds. Comfortably hard, like threshold runs or marathon pace for many runners.
- Zone 3 (hard): above the second threshold. Intervals you can hold for minutes, not hours.
“Easy” on this page means zone 1. If you use a five-zone system on a watch, the labels differ. See how to set your heart-rate zones.
What 119,452 marathoners did
The largest data set comes from training logs of 119,452 marathoners covering 151,813 marathons (Muniz-Pumares and colleagues, 2025). The researchers looked at the 16 weeks before each race.
The fastest runners, who finished in 2:00–2:30, ran more than three times the volume of slower runners. Yet their total time in zones 2 and 3 was about the same as everyone else’s. Almost all the extra running was in zone 1. More than 80% of these fast runners trained in a pyramidal pattern: lots of easy, some steady, a little hard. (For more on that pattern, see polarized vs pyramidal training.)
Two other large data sets agree.
- In a Polar data set of about 14,000 runners, faster runners trained at lower relative intensities. The benefit of adding more volume and intensity together levelled off at some point (Emig & Peltonen, 2020).
- In a Runtastic analysis of 6,771 race finishers, the runners who improved most had the most low-intensity volume. Their high-intensity running was 5% of volume or less (Zrenner and colleagues, 2021).
Both of these studies have company ties. The first is affiliated with Polar, and the second has Adidas/Runtastic co-authors. That doesn’t make them wrong, but it is worth knowing.
What the best runners do
World-class distance runners keep at least 80% of their running distance at low intensity, all year round. Marathoners lean more pyramidal than 1500 m runners (Haugen and colleagues, 2022; Casado and colleagues, 2022).
What the trials say
Training trials are short and use shorter races, but they point the same way. A review of 20 studies found that at least 70% low-intensity training was the common thread of approaches that worked. Both polarized and pyramidal plans did better than plans built mostly around threshold work (Campos and colleagues, 2022).
So the evidence doesn’t give one magic percentage. It gives a floor of about 70% and a clear pattern: more easy running, not more hard running, separates faster runners from slower ones.
A practical target
For a recreational marathoner, a reasonable split by running time is:
- Zone 1: 75–85%
- Zone 2: 10–20%
- Zone 3: 3–8%
These bands are our suggestion, built from the studies above. No trial has tested these exact numbers. Another suggestion: keep the combined zone 2 and zone 3 time at roughly 60–120 minutes a week, and add any extra time as easy running. That mirrors what the fastest marathoners did, but it hasn’t been tested as a rule either.
Keep volume increases safe. Raising your weekly running is the point, but a single run that jumps well past your recent longest run is linked to more injuries. See the long-run rule.
How you count changes the answer
The same training can look different depending on how you count it. Applied to the same world-class training logs, one counting method gave a zone-3 share of 4.1% and another gave 6.6% (Kenneally and colleagues, 2021). Counting by time or by sessions also gives different numbers.
A week with two interval sessions out of six runs is “33% hard” by sessions. By time, it may be under 10% hard, because the warm-up, recoveries and cool-down are easy running. Pick one method, ideally time, and use it consistently.
What we don’t know
- The big data sets are observational. They can’t separate cause from selection. Fitter runners may simply choose to run more, rather than running more making them fitter.
- No trial has tested a specific easy percentage for marathon performance. The trials use shorter time trials and last weeks, not a full marathon build.
- The ideal share for an individual is unknown. The 75–85% band is a reasonable default, not a proven optimum.
- Women are under-represented in the trial data, at about 15% of participants.
Sources
- Muniz-Pumares et al. (2025). Training characteristics of 119,452 marathon runners.
- Emig & Peltonen (2020). Training intensity and performance in about 14,000 Polar users.
- Zrenner et al. (2021). Training response in 6,771 marathon finishers using Runtastic.
- Haugen et al. (2022). The training characteristics of world-class distance runners.
- Casado et al. (2022). Training intensity distribution in elite middle- and long-distance runners.
- Campos et al. (2022). Review of training intensity distribution in endurance athletes.
- Kenneally et al. (2021). Training intensity distribution analysis by race pace vs physiological approach.
Last reviewed 13 November 2026.