You skipped a week of running. Your race is still on the calendar. Here’s the short answer: don’t cram, restart at about 70% of planned volume, and protect your long run.
A week off costs less than the panic suggests. A big Strava study found that a 7 to 13 day stretch with no training at all made marathons about 4.25% slower. Longer gaps cost more.
What a missed week of marathon training costs your race time
Researchers at University College Dublin compared 56,735 pairs of marathons from 43,933 runners. Each runner served as their own control. They compared a race after a long zero-training gap with one after no gap longer than 6 days (Feely et al., 2023).
For a 4:00 marathoner, 4.25% is about 10 minutes. A 14-day gap is about 14 minutes. A 28-day gap is about 19.
In short: one lost week stings. One lost month hurts a lot.
Here’s what most articles skip. “Gap” meant days with zero logged activity of any kind. If you cross-trained, swam, or walked hard, you weren’t measured. The study also can’t tell injury from illness from a busy month, and it only counts runners who finished.
A summary of the paper also reports that faster runners and men lost more, about 5% versus 3.4% for slower runners and women. Treat that split as secondary until you read the paper yourself.
Your own penalty is probably smaller if you stayed active.
What actually happens to your fitness
Picture fitness as a bank account, not a sandcastle. Skipping a week is a small withdrawal. It isn’t a washout.
In a classic 1984 study, trained athletes stopped all training. Their VO2max fell about 7% in the first 12 to 21 days. It then settled about 16% lower by day 56 (Coyle et al.). VO2max is the ceiling on how much oxygen your body can use. Only seven athletes took part, so treat the numbers as a rough guide.
A 2022 meta-analysis of 21 studies found an average VO2max drop of about 3.9% in the first 30 days off, and about 9.4% after that (Zheng et al.). Individual results varied a lot.
Why so small early? Part of the first loss is blood plasma volume. It drops fast and comes back fast. Muscle enzymes that burn fuel fall more slowly, with a half-time near 12 days. And the tiny blood vessels feeding your muscles didn’t shrink at all in that 1984 study.
Translation: after one week, most of what you lost is the quick-to-return kind.
Your fitness has a short memory for rest and a long memory for work.
Why cramming makes it worse
You’ll want to run the missed week’s miles next week. Don’t.
A missed week drops your recent load. Jump back to a double week and your body sees a sudden spike. Researchers have tied big weekly jumps to injury. In 874 novice runners, those who raised weekly distance by more than 30% had more distance-related injuries than those who stayed under 10% (Nielsen et al., 2014).
Two honest caveats. Those were novices, not marathoners. And the 10% figure was a comparison group, not a proven safe line.
You may have also seen the rule that a load ratio above 1.5 predicts injury. A 2020 paper argues that ratio has conceptual and statistical flaws and shouldn’t be used to prescribe training (Impellizzeri et al.). Use it as a rough smell test only.
The practical point survives. Your body adapts to what you did recently, not to what the plan says. If you want to see why simple ramp rules still break down, read why the 10% weekly mileage rule is wrong.
Make-up miles are a loan with interest.
The adjustment plan by weeks to race
The timing bands below match the Strava study’s windows. That study found gaps 3 to 7 weeks out cost about 5.2%, versus about 4.4% for gaps 8 to 12 weeks out. The study didn’t test the “more than 12” and “under 3” rows, so those are coaching bands. The re-entry numbers and priorities in every row are coaching consensus. No trial has tested them.
| Weeks to race | Re-entry volume | Don’t make up | Protect first | Evidence |
|---|---|---|---|---|
| More than 12 | ~70% after 1 week off | The missed miles | Next easy build week; slide the plan or repeat a week | Timing: coaching band. Plan: coaching |
| 8 to 12 | ~70-85%, then rejoin | The missed week | Your next long run | Timing: study window. Plan: coaching |
| 3 to 7 | ~70% (1 wk) or ~50% (2 wk) | Easy volume; drop it | One long run and one quality session a week | Timing: study window. Plan: coaching |
| Under 3 (taper) | Hold steady | Anything | A few short race-pace touches, sleep | Coaching band |
The 3 to 7 week window is where you have the most to lose and the least room. Don’t slide the race. Cut easy junk miles first. Aim to regain your peak long run only if you can still get two full weeks of taper.
Inside 3 weeks, a missed week is mostly harmless. Taper already lowers load. Trust the fitness you built.
If you can only do part of a week
Rank your sessions. Coaches usually order them like this:
- The long run.
- Quality sessions, such as tempo or intervals.
- Marathon-pace work.
- Easy runs.
Most coaches say finishing about 80% of your plan is enough if the long runs and quality days are in it. That’s coaching opinion, not study data. A 2025 look at Boston Marathon runners also hinted that fewer, better sessions can beat piling on volume. It’s observational, so don’t lean on it.
Illness, travel, or injury: different rules
The cause changes the restart.
Travel or a busy week. You’re fine. Restart at about 70% with easy running, then rejoin the plan.
Illness. Skip training with a fever or chest symptoms. Rest from one day up to about a week after a febrile illness is typical advice. The old “neck check” rule (train if symptoms sit above the neck) was called out in a 2022 review as unsupported and possibly hazardous (Sports Medicine, 2022). Return at about half your usual volume once symptom-free. No hard sessions for the first few days.
Injury. The Strava data doesn’t cover this. Get cleared first. Cross-train to hold your aerobic base, then return with a walk/run progression. See a clinician if pain lingers.
A real-world example
Take a runner I’ll call Dan. He’s 41, targets 3:45, and peaks near 40 miles a week. Ten weeks out, flu took him down for 8 days with zero running.
His old plan said 40 miles and a 16-mile long run that week. He wanted to run 56 miles to catch up. Instead, he rested until the fever cleared, then ran 28 miles (70%) with all easy effort. The next week he did 34 miles with a 12-mile long run and one tempo session. By week 7 he was back at 40 miles with a 16-mile long run.
He lost about 8 days. He didn’t lose the race. He’s inside the study’s 7 to 13 day band, where the average penalty was about 4.25%, and he had 10 weeks to rebuild.
Let the plan do the math
You don’t have to run this table in your head. AthleteOS sees the gap through workout import. Its AI coach then rebuilds your training plan from your own recent load, not a generic chart. It re-enters at a lower volume, protects the long run, and drops low-value easy miles instead of piling on make-up work. It tells you why it changed each workout.
If you also track base fitness, how Zone 2 builds the aerobic base explains what the restart weeks are protecting. And what the drift ratio tells you about your base is a quick check that your fitness came back, not just your mileage.
Missed a week? Skip it and keep going.