← All articles
Running

How Long Should Your Long Run Actually Be?

The long run is the most over-prescribed session in distance running. Why time on feet beats distance, what percentage of weekly volume it should be, and the point past which a longer run stops buying anything.

OnlyGains Editorial··5 min read
The short version
  • Measure the long run in time, not distance. A slower runner doing 32 km is doing a different session.
  • Keep it to roughly 25 to 30% of your weekly volume.
  • Around 2.5 to 3 hours is where the cost starts outrunning the benefit for most people.
  • Run it easy. A long run at moderate effort is a hard session you did not plan for.
  • Two medium-long runs a week often beat one enormous one.

Ask most recreational runners what their long run should be and the answer arrives in kilometres — 20, 25, 32. Ask what it is for and the answers get vaguer.

That gap is why the long run is simultaneously the most respected and most poorly executed session in distance running.

What it is actually for

Three adaptations, none of which require heroic distance.

Aerobic development. Sustained submaximal work drives mitochondrial density, capillarisation and increased plasma volume. These respond to accumulated duration.

Fat oxidation and glycogen sparing. Extended running trains your ability to use fat at a given intensity, sparing the glycogen that runs out around the two-and-a-half-hour mark.

Musculoskeletal durability. Tissue tolerance for repeated impact, which is the adaptation that takes longest and matters most for staying uninjured.

Note what is not on that list: mental toughness. It is a real side benefit, but it is not a physiological reason to run four hours, and it is the justification people reach for when the physiological one has run out.

Measure it in time, not distance

This is the most useful change most runners can make.

Two people run 32 km. One takes 2 hours 40 minutes; the other takes 3 hours 50 minutes. They have done fundamentally different sessions — different fuel demands, different muscle damage, different recovery cost, different injury risk.

Distance-based prescriptions systematically overload slower runners, who are usually the ones with the least tissue tolerance to spare.

So prescribe time. "Two hours easy" is a session that means the same thing to everyone.

The numbers

GoalLong run durationCeiling
General fitness45–75 min
10K60–90 min
Half marathon75–120 min~2 h
Marathon2–3 h3 h
Ultra3 h+, with back-to-backsVaries

Roughly 25 to 30% of weekly volume. If you run 50 km a week, the long run is 13 to 15 km. If your long run is 40% or more of your week, you have a distribution problem: too much stress in one session, not enough base across the others.

This is the single most common structural error in amateur marathon training — a 30 km long run on top of three 6 km midweek runs. That week is 50% long run, and it will hurt somebody.

The ceiling

Somewhere around two and a half to three hours, the trade turns.

Past that point:

  • Muscle damage rises steeply, and the next two to three days of training degrade
  • Immune function is transiently suppressed
  • Injury risk climbs, particularly in the final third when form deteriorates
  • The additional aerobic benefit is marginal

Which produces the counterintuitive conclusion for slower marathoners: if 32 km would take you four hours, do not run 32 km. Cap at three hours, and build your weekly volume instead. You will arrive at the start line fitter and considerably less broken.

Elite marathoners run 35 km+ long runs, but they cover it in around two hours. They are not doing a longer session than you — often they are doing a shorter one.

Pace

Comfortably conversational. Typically 60 to 90 seconds per kilometre slower than marathon pace.

The most common error is running the long run at moderate effort throughout, which converts a base-building session into an unplanned hard session and compromises the next two days. This is the same grey-zone failure described in why your 5K time is stuck, applied to the one session where it does the most damage.

The exception: marathon-pace segments in the final third during a specific build. Finishing a long run with 6 to 10 km at goal pace is one of the most valuable sessions in marathon training, precisely because it rehearses running goal pace on tired legs. That is deliberate, planned, and confined to a specific block — not a general licence to run the long run hard.

The alternative nobody considers

Two medium-long runs instead of one very long one.

Instead of one 3-hour run, do a 1:45 and a 1:30 in the same week. You get:

  • Similar total aerobic volume
  • Substantially less muscle damage
  • Faster recovery, so the rest of the week stays intact
  • Lower injury risk

When this is the better choice: if you are injury-prone, over 40, in a calorie deficit, short on sleep, or your long run currently exceeds 30% of your weekly volume.

When the single long run is better: marathon and ultra specificity, particularly in the final 8 to 10 weeks, where the continuous duration is the point.

Fuelling it

Anything past 90 minutes should be fuelled, and this is also your opportunity to train the gut rather than just the legs.

Take 30 to 60 g of carbohydrate an hour on long runs, even when you feel you do not need it. Practising this is the entire mechanism described in gut training, and the long run is the only place to do it.

Progression

Add roughly 10 to 15 minutes every two weeks, not every week.

Every fourth week, cut it by a third. The long run should follow the same down-week rhythm as everything else — see deload weeks.

Do not increase duration and intensity in the same week. Pick one.

The summary

Time, not distance. Roughly a quarter to a third of your week. Comfortably easy, except for deliberate goal-pace segments in a specific block. Capped around three hours regardless of what the plan says.

And if the long run is the only substantial run in your week, the problem is not your long run. It is the four days around it.

Questions people ask

Short answers to the things this article gets asked most.

How long should a long run be?

For most recreational runners, 90 minutes to 2.5 hours covers nearly all the benefit, and 3 hours is a sensible ceiling. Think in time rather than distance — a 5-hour marathoner running 32 km is on their feet for over three and a half hours, which is a completely different session from a 3-hour marathoner covering the same distance.

Should you run 20 miles before a marathon?

It is a useful psychological milestone and a common one, but it is not a physiological requirement. What matters is time on feet and total weekly volume. If 32 km takes you well over three hours, you are better served by capping the run at around three hours and building your weekly volume instead.

What percentage of weekly mileage should the long run be?

Roughly 25 to 30%. If your long run is 40% or more of your week, the week is unbalanced — you are taking most of your training stress in a single session, which raises injury risk and leaves the rest of the week too light to build a base.

How slow should a long run be?

Comfortably conversational, typically 60 to 90 seconds per kilometre slower than marathon pace. The exception is deliberate marathon-pace segments late in the run during a specific build. A long run drifting into moderate effort throughout becomes an unplanned hard session and compromises the rest of the week.

Can you run too long on a long run?

Yes. Past roughly three hours the muscle damage, immune suppression and recovery cost rise faster than the aerobic benefit, and the following days of training suffer. Very long runs also carry disproportionate injury risk, particularly for runners whose weekly volume does not support them.

long runmarathon trainingtime on feetweekly volumeendurancerunning plan

Keep reading