Written by Tommo, Verus founder · Last verified August 2026 · 9 min read

Running Guide

How Long Does It Take to Train for a Marathon?

Quick Answer

For someone who can already run comfortably for 30 minutes, a marathon takes 16 to 20 weeks of training. From a standing start, allow 9 to 12 months: roughly six months to build to a comfortable 10K and a 20-mile week first, then the 16-to-20-week block on top. The limit is rarely fitness — it is how fast your tendons, bones and joints adapt.

This question has one honest answer and one popular answer, and they are not the same. The popular answer is sixteen weeks, because that is the length of most published plans. The honest answer is that sixteen weeks is how long the final block takes, and the length of everything before it depends entirely on where you are starting from.

So this guide answers it from four different starting points, explains what actually sets the limit, and shows you how to work backwards from a race date to the week you need to begin.

Or skip straight to the number: pick your starting point and get the weeks needed, plus a start-by date for your race.

The four starting points

Where you are nowTime to the start lineWhat the time is for
Not currently running9-12 monthsSix months or so building to a comfortable 10K and a consistent 20-mile week, then a full 16-20 week marathon block on top.
Running 30 minutes comfortably6-8 monthsTwo to three months raising weekly volume and getting a long run to 10 miles, then the 16-20 week block.
Running 15-20 miles a week, can run a 10K16-20 weeksStraight into a marathon block. This is the runner every published plan is written for.
Running 30+ miles a week with a recent half marathon12-16 weeksYou already have the base. The block is about marathon-specific long runs and race-pace work.

The row that matters most is the third one, because almost every training plan you can download assumes you are in it. A plan that opens with a 10-mile week and a 6-mile long run is not a beginner plan; it is an intermediate plan with a gentle first fortnight. If you cannot currently run 6 miles, week one of that plan is already a hard session.

The honest test

Can you run for an hour, today, without stopping and without it being an event? If yes, you are ready to start a 16-week block. If no, the sixteen weeks starts after you can — and the time it takes to get there is the part nobody counts.

What actually sets the limit

Cardiovascular fitness improves quickly. Within six to eight weeks of consistent running, most people are noticeably fitter — heart rate at a given pace drops, easy runs stop feeling like work, and the temptation to compress the timeline becomes strong.

The structures that carry you — tendons, ligaments, bone, connective tissue — adapt far more slowly, on a timescale of months rather than weeks. That mismatch is the single most common cause of marathon training going wrong: your engine outpaces your chassis, you feel great, you add volume to match how you feel, and something in a shin or an achilles or a hip objects.

This is why the answer to "how long does it take" is not a fitness question. Nearly everyone who fails to make a marathon start line healthy did so by treating it as one.

  • Increase weekly volume gradually and take a lower-volume week roughly every fourth week. The exact percentage matters less than the principle of not adding week on week indefinitely.
  • The long run is the session that builds the marathon and the session most likely to injure you. Extend it in small steps and not every single week.
  • Most of your running should be genuinely easy — conversational, and slower than feels dignified. This is the part people skip, and it is the part that lets the volume happen at all.
  • Consistency over months beats intensity over weeks. Four steady months will get you further than six brilliant weeks and a stress fracture.

Working backwards from a race date

The practical version of all of this is a subtraction. Take your race date, subtract the block, subtract the base-building, and you have the week you start — and, more usefully, an honest answer about whether the race you are looking at is the right one.

Race dateBlock startsIf you need base first, start
A spring marathon in late AprilEarly JanuaryThe previous summer
An autumn marathon in mid-OctoberLate May / early JuneThe previous winter

Most UK marathons fall into two clusters, which makes this easier than it sounds. The spring cluster runs from early April to the end of May — Brighton, London, Manchester, Milton Keynes, Edinburgh — and the autumn cluster sits in October, with Chester, Yorkshire and Abingdon within a week of each other. Pick a cluster, count backwards, and you have your start date.

If the subtraction produces a date in the past, that is useful information rather than a failure. It means you want the next cluster, and the extra six months will produce a better race and a healthier build.

Before you commit to a date, see what your current fitness actually converts to over 26.2 miles.

What the 16 to 20 weeks are actually for

A marathon block is not simply more of what you were already doing. It has a shape, and understanding the shape tells you why it cannot be meaningfully compressed.

PhaseRough lengthWhat it does
Base4-6 weeksRaise weekly volume, get the long run moving, add strides. No heroics.
Build6-8 weeksThe longest long runs, threshold work, and marathon-pace running inside them.
Peak2-3 weeksThe hardest fortnight of the block, and where a well-executed plan feels genuinely difficult.
Taper2-3 weeksVolume drops 40-60%, intensity largely stays. This is where the training becomes fitness.

Notice that the taper is two to three weeks of the sixteen. That is not slack in the plan you can reclaim — cutting it is one of the more reliable ways to arrive at the start line tired, and the marathon taper guide covers why in detail.

Can you do it faster than that?

People do. Twelve-week plans exist, people complete marathons off eight weeks of training, and someone you know finished one on almost nothing and will tell you about it. All of that is true and none of it changes the advice.

What a compressed timeline actually buys you is a higher chance of one of three outcomes: an injury before the start line, a race that becomes a walk somewhere after twenty miles, or a recovery that takes considerably longer than the training did. If the date is fixed and the timeline is short, the sensible adjustment is the goal rather than the schedule — go in aiming to finish comfortably rather than to hit a time, and treat it as the first marathon rather than the only one.

The one exception worth naming

If you have run marathons before and have been consistently running through the year, 12 weeks is genuinely enough, because you never lost the base. The timeline is short for people building from nothing, not for people returning to something.

How many hours a week does it take?

The other half of the question, and the one that decides whether the plan survives contact with your life. Rough guide, for the block itself rather than the base-building:

GoalDays runningTypical weekly hours at peak
Finish comfortably3-44-6 hours, with the long run taking half of it
Run a specific time4-56-8 hours
Chase a personal best5-68-10 hours or more

The long run dominates all three, and it dominates increasingly as the block goes on. A three-hour long run plus fuelling, changing and being fit for anything afterwards is most of a weekend morning, and it happens most weekends for four months. That is the commitment people underestimate — not the hours, but which hours.

Put a date on it

The single largest difference between people who complete a marathon build and people who intend to is whether there is a date. An entry paid for and a start line in the calendar converts a vague ambition into a series of Tuesdays that have to happen.

This is also, for what it is worth, the whole idea behind a Verus profile: you set where you are starting from and the date you are aiming at, and the journey between them is the thing on the page. It is a public commitment rather than a private intention, and public commitments survive February better.

Once you have a date, get the five paces the block should actually be run at, from a race result you already have.

Frequently Asked Questions

How long does it take to train for a marathon from scratch?

Nine to twelve months if you are not currently running. Roughly six months to build to a comfortable 10K and a consistent 20-mile week, then a full 16 to 20 week marathon block on top of that. The limit is not fitness — tendons, bone and connective tissue adapt over months, while cardiovascular fitness improves in weeks.

Is 16 weeks enough to train for a marathon?

Yes, if you can already run for about an hour without stopping and are covering 15 to 20 miles a week. That is the runner almost every published 16-week plan is written for. If you cannot, the sixteen weeks starts after you can — and the time to get there is the part most people forget to count.

Can you train for a marathon in 12 weeks?

If you have run marathons before and have kept running consistently through the year, yes — you never lost the base, and twelve weeks is enough to sharpen it. Building from nothing in twelve weeks mostly buys a higher chance of injury, a walk after twenty miles, or a long recovery. With a fixed date and a short runway, change the goal rather than the schedule.

How many hours a week is marathon training?

Four to six hours a week at peak to finish comfortably, six to eight to run a specific time, and eight to ten or more to chase a personal best. The long run is half of it and grows through the block, so the real commitment is most of a weekend morning, most weekends, for four months.

When should I start training for a spring marathon?

Early January for a late-April race, if you already have a base. Most UK marathons fall into two clusters — spring from early April to the end of May, and autumn in October — so pick a cluster and count backwards. If the subtraction lands in the past, that is a signal to target the next cluster rather than to compress the plan.