Does the 10% rule work?
Not as an injury guard. In a randomized trial, novices who followed the 10% rule got injured just as often as those who built up faster, and a study of 5,205 runners found no link between week-to-week mileage changes and injury. What did matter was single runs that were much longer than anything recent.
What to do
- Stop using the weekly 10% rule as your safety check. It has been tested and it didn't work.
- Check each run instead. Keep every run within 10% of your longest run in the last 30 days.
- Don't worry about chasing a 'sweet spot' load ratio from your watch. That measure is heavily disputed.
- If you are new to running, avoid very large jumps over a couple of weeks, such as growing your volume by more than 30%. That is a cautious suggestion, not a proven limit.
- After a break, rebuild from what you have actually run recently, not from where you were before.
Where the rule comes from
The 10% rule says you should never increase your weekly mileage by more than 10% from one week to the next. It is simple, it sounds careful, and it appears in countless plans and apps. But it is running folklore, and when researchers tested it, it did not hold up.
The randomized trial
The most direct test is a randomized trial in 532 novice runners (Buist and colleagues, 2008). One group followed a graded programme built on the 10% rule. The other followed a standard programme that built up faster.
Injury rates were the same: 20.8% in the 10% group against 20.3% in the faster group. Building up more slowly did not protect anyone.
The observational studies
Three cohorts have looked at weekly progression in real runners.
- 874 novice runners (Nielsen and colleagues, 2014). Runners who raised their weekly distance by more than 30% showed only a weak trend towards more distance-related injuries, and it was not statistically significant.
- 5,205 Garmin users, 588,071 training sessions (Frandsen and colleagues, 2025). Week-to-week changes in total distance showed no link with injury. What did predict overuse injury was a single run more than 10% longer than the longest run of the previous 30 days.
- 735 runners preparing for the 2019 New York City Marathon (summarized on the BJSM blog). Breaking the 10% rule was not linked to injury. More days with a high acute:chronic workload ratio (explained below) were.
So the best data agree on one point: the weekly 10% line itself doesn’t separate safe training from risky training.
What about the acute:chronic workload ratio?
The acute:chronic workload ratio (ACWR) compares your last week of training with your average over the last four weeks. It is often presented as a smarter version of the 10% rule, with a “sweet spot” between 0.8 and 1.3.
It has serious problems. A methodological critique showed that the way the ratio is calculated can create links with injury even in simulated data where no real link exists (Impellizzeri and colleagues, 2020). The evidence in runners is also mixed. In the New York cohort, high-ratio days were linked to more injuries. In the much larger Garmin cohort, a higher ratio was linked to fewer injuries.
The reasonable reading is that sudden jumps in training do matter, but the ACWR is a noisy way to spot them. We would treat it, at most, as a rough consistency indicator, not as a gate.
What to use instead
The single-run rule from the Garmin study catches sudden jumps more directly. Before each run, find your longest run in the last 30 days and multiply it by 1.1. That is your ceiling. We explain the study, the injury figures and how to build a long run with it in the long-run rule that actually prevents injuries.
One practical point carries over from the weekly-rule debate. Novices may still want some guard on fast growth. The 30%-plus trend in the Nielsen study was not significant, but it pointed the expected way. A soft warning when volume grows by more than about 30% over two weeks is our cautious suggestion for new runners, not a tested rule.
Injury risk also depends on who you are. New runners are injured more than twice as often per hour as experienced recreational runners (Videbæk and colleagues, 2015), and a previous running injury is the strongest known risk factor (van Poppel and colleagues, 2021). Those runners have the most reason to be careful, whatever rule they use.
What we don’t know
- The trial was in novices training for a short race. It doesn’t tell us about experienced runners building towards a marathon.
- The cohorts are observational. They show which patterns were linked to injury, not proof of cause.
- No study has tested a better weekly number. The finding is that the 10% line doesn’t help, not that some other weekly percentage does.
- The New York cohort is known to us only through a blog summary, so its details are less certain than the others.
- Load is more than distance. Speed, hills, surfaces and recovery also matter and are not captured by weekly mileage.
Sources
- Buist et al. (2008). Randomized trial of a graded training programme (the 10% rule) in novice runners.
- Nielsen et al. (2014). Weekly progression in running distance and injury in 874 novice runners.
- Frandsen et al. (2025). Garmin-RUNSAFE: single-session running distance spikes and injury in 5,205 runners.
- BJSM blog (2023). Summary of a 2019 New York City Marathon cohort on training patterns and injury.
- Impellizzeri et al. (2020). Critique of the acute:chronic workload ratio.
- Videbæk et al. (2015). Injury incidence per 1,000 hours of running in different runner groups.
- van Poppel et al. (2021). Risk factors for running injuries.
Last reviewed 13 November 2026.