How should you recover after a marathon?
You don't need a week of total rest. In a trial, easy running or elliptical work from 48 hours after a marathon did not slow muscle-damage recovery. Keep the first two weeks easy and short, and build back over weeks three and four; that later timeline is convention, not tested.
What to do
- Rest or walk for the first two days.
- From 48 hours, an easy 30–40-minute run or elliptical session every couple of days is fine if you feel like it.
- Skip hard sessions and long runs for the first two weeks, then return gradually over weeks three and four. This part is convention.
- Within 1–2 hours of finishing, eat about 0.3 g of protein per kg of body weight (20–40 g) and 1–1.2 g of carbohydrate per kg.
- If you are sore in a way that changes how you run, or have pain at one spot on a bone, pain when hopping or pain at night, stop and see a doctor or physio.
Rest or move?
A common piece of advice is to take the whole week after a marathon off. The best evidence says that is not necessary.
In a trial, runners who did easy running or elliptical work from 48 hours after a marathon recovered from muscle damage just as well as those who rested (Martínez-Navarro and colleagues, 2021). Light activity did not slow things down.
That does not mean you have to run. It means a short, easy session from day three is a reasonable choice if you want one, and total rest is optional rather than required.
A return plan
This comes from our source report’s rule set (RC-1). Only the first part is backed by the trial; the rest is coaching convention.
| Period | What to do | Evidence |
|---|---|---|
| Days 0–2 | Rest or walk. | Convention |
| From 48 hours | Optional easy run or elliptical, 30–40 minutes, about every second day. | B (one trial) |
| Weeks 1–2 | No hard sessions, no long runs. | Convention |
| Weeks 3–4 | Gradual return to normal training. | Convention |
One point is our own suggestion. The long-run injury rule compares each run with your longest run in the past 30 days. Straight after a marathon, that longest run is the race itself, so the rule will not hold you back. Use the timeline above, not the spike rule, to pace your return in the first month.
If you can’t run
If something hurts and running is off the table, other activities can keep your fitness ticking over. Deep-water running (running motions in water too deep to touch the bottom) maintained cardiorespiratory fitness about as well as land training in a review of studies (deep-water running review). Elliptical or cycling are also options. The rule-set default is to match the duration and effort of the runs you would have done. That substitution rule is graded C.
Eating after the race
- About 30 g of protein after exercise maximized muscle-building in endurance-trained men (Churchward-Venne and colleagues, 2020).
- The practical rule built on this (FU-11, grade B) is about 0.3 g of protein per kg of body weight (20–40 g) plus 1–1.2 g of carbohydrate per kg, within 1–2 hours of a long run or race.
As an example: a 70 kg runner would aim for about 21 g of protein and 70–84 g of carbohydrate in that window.
A strict 30-minute “window” for protein, often promoted to active women on social media, lacks robust research support (Matkin-Hussey and colleagues, 2026). The rule above uses 1–2 hours. For fuelling during the race itself, see how many carbs per hour.
Sleep
The research linking sleep to injury is weak and comes mainly from adolescents (Hatia and colleagues, 2024). A convention-based default in our source report is: if you sleep less than 7 hours on three or more nights in a week, swap your next hard session for an easy run. Treat that as a sensible habit, not a proven rule.
Red flags
Watch for these warning signs. The thresholds come from expert practice rather than trials (RC-5):
- Pain of 3 out of 10 or more that changes how you run: cut back or stop.
- Pain at one spot on a bone, pain when hopping, or pain at night: stop running and see a doctor or physio.
Never try to run through these. Masters runners report more fatigue and soreness, though evidence that recovery itself slows with age is limited; see masters runners.
What we don’t know
- The 48-hour finding comes from one trial that measured muscle damage, not later race performance.
- The week-by-week return plan is convention. No study has compared different return timelines.
- The protein figure comes from men. Whether women need different amounts is not settled by this evidence.
- Sleep evidence is weak and mostly from young athletes.
Sources
- Martínez-Navarro et al. (2021). Trial of easy running or elliptical work after a marathon.
- Deep-water running review. Review of deep-water running for maintaining fitness.
- Churchward-Venne et al. (2020). Protein dose and muscle-building after exercise in endurance-trained men.
- Matkin-Hussey et al. (2026). Review of nutrition claims aimed at active women on social media.
- Hatia et al. (2024). Review of sleep and injury in athletes.
- Borges et al. Review of recovery in masters athletes.
Last reviewed 13 November 2026.