How to increase your running mileage safely

Updated August 26, 2026 · By the Runked team

Raise mileage too fast and injuries follow. Hold it too flat and fitness stalls. The advice collapses into a single number, the 10% rule, but the research behind that number is narrower than the rule's reputation suggests, and it splits depending on whether you apply it to a single run or a whole week.

What two studies on mileage growth actually measured

Buist and colleagues, publishing in the American Journal of Sports Medicine in 2008, put 532 novice runners on one of two programs, both training toward the same 4-mile event: a standard 8-week program or a 13-week program that raised mileage by roughly 10% a week. The trial found injury rates almost identical between groups, 20.8% versus 20.3%. A 2025 cohort study of 5,205 runners by Frandsen and colleagues, published in the British Journal of Sports Medicine (Frandsen and colleagues), defined three exposures: whether a single session exceeded the runner's own longest run of the previous 30 days by more than 10%, whether the week's total mileage exceeded the week before by more than 10%, and the acute:chronic workload ratio, one week's distance against the preceding three. Overuse injury rates rose significantly on the first measure, with an adjusted hazard rate ratio of 1.64 for jumps in the 10-to-30% band. The week-to-week ratio, the measure Buist's trial tested directly, showed no relationship at all, and the acute:chronic ratio ran the other way: a negative dose-response, where the bigger spikes went with lower injury rates rather than higher.

The acute:chronic workload ratio: a second way to frame it

The 10% figures above describe one ratio: this week, or this run, against the recent past. A related metric, the acute:chronic workload ratio (ACWR), formalizes that same comparison for training load broadly, beyond weekly mileage alone. Tim Gabbett's 2016 review in the British Journal of Sports Medicine compares a runner's recent load, roughly the last 7 days, against their longer-term average load, roughly the last 4 to 6 weeks, as a ratio. That review associates a ratio of roughly 0.8 to 1.3 with lower injury risk, with risk climbing as the ratio rises above roughly 1.5.

The ratio has published critics. Impellizzeri and colleagues, in a 2020 paper in the International Journal of Sports Physiology and Performance, argued that the ratio's acute and chronic components share overlapping data by construction, a statistical issue known as mathematical coupling, and concluded that "the statistical properties of the ratio make the ACWR an inaccurate metric." They found no study has tested whether deliberately changing a runner's ACWR changes their injury risk, and described using the ratio to guide practical training decisions as an area where "manipulating ACWR in practical settings in order to change injury rates remains a conjecture." Read next to Gabbett's review, the evidence on ACWR is mixed: an association between high ratios and injury shows up in observational data, and the statistical basis for treating the ratio itself as a lever remains disputed.

Daniels' cap on any single long run

A separate question is how large one run should be relative to the week around it. Exercise physiologist Jack Daniels, in Daniels' Running Formula, ties the cap to weekly volume rather than one flat number: under about 40 weekly miles (64km), no single long run should exceed 30% of that week's total; at 40 or more weekly miles, the cap drops to whichever is smaller, 25% of the week's mileage or 150 minutes. A runner logging 20 weekly miles falls in the lower tier, so the cap is 30% (about 6 miles), regardless of how good that run feels partway through.

A build/step-back structure that applies these findings

The table below is Runked's own illustration, built by applying two of the figures above across a single four-week block: Daniels' tiered cap on the long run, and the single-run growth threshold Frandsen's cohort associated with rising injury risk. It starts from an example base of 20 miles a week; scale the numbers to your own base mileage.

Week 5 would restart the build from around 18 to 20 miles rather than continuing from week 3's peak. The step-back week keeps three weeks of consecutive growth from turning into four or five, which is the point at which both the single-run threshold above and a rising ACWR become harder to avoid.

Common mistakes when increasing mileage

• Applying the 10% rule only to the week's total. Frandsen's cohort tied injury risk to the single run's jump past the 30-day longest run; the week-to-week ratio showed no relationship in that same study.
• Skipping the step-back week. Stacking a fourth build week on top of three removes the recovery the previous three weeks were building toward.
• Raising mileage and intensity at the same time. Adding tempo runs or intervals in the same week as a mileage increase changes two variables at once.
• Treating ACWR as a precise diagnosis. Impellizzeri and colleagues' critique targets the ratio's statistical properties specifically, so one number crossing 1.3 isn't proof of anything on its own.

What to do next

Runked's 5K and marathon training plans build weekly mileage through six phases, Foundation, Early Quality, Transition, Race Prep, Taper and Recovery, instead of a flat percentage. The training load guide covers the ACWR concept from the runner's-eye view, as Fitness, Fatigue and Load Ratio. And a weekly mileage target by level is a reasonable starting base before applying any of the growth rules above. Compare this week to your own last few weeks, not to a flat rule or someone else's plan: in Runked the rank starts everyone at Bronze and only moves on your own runs.

Frequently asked questions