OUTRN

How to Train for Your First Ultra Marathon

Training | Published 2026-10-03 | By OUTRN Team

The OUTRN team breaks down what the research says about training, fuelling and tapering for a first ultra, and where the evidence runs out...

Ask five ultra runners how to train for a first 50km and you'll get six plans. Ask the research and you get something more awkward: a handful of strong answers on fuelling and tapering, and a lot of polite silence on the questions you actually care about (how many kilometres? how many weeks?).

Here is the short answer: build your longest run gradually, practise your fuelling until your gut tolerates it, drink to thirst, add strength work and taper for up to three weeks. The research backs those habits best. It cannot yet tell you how many kilometres or weeks a first ultra needs.

This guide to ultra marathon training is for someone who already runs 10km comfortably and is wondering what a first ultra takes. We went through the studies, opened the abstracts, and kept only what we could verify. Where the evidence is thin we say so, and where we're offering a coach's convention instead of a finding we label it. Let's start with what counts as an ultra...

What counts as an ultra marathon?

An ultramarathon is any distance longer than the traditional 42.195km marathon, according to World Athletics (2009 wording). In practice, a first ultra is usually a 50km trail race.

As a worked example, the UTA50 at Ultra-Trail Australia in 2027 is 51km with 2,400m of climbing and a 14 hour 30 minute limit, and it finishes by climbing the 951 Furber Steps. General entry is currently sold out: our Ultra-Trail Australia guide has the details.

How hard is a first ultra to finish?

Honestly? Nobody can give you a clean number. We found no peer-reviewed finish or did-not-finish rate for 50km, 80km or 100km trail races.

The closest data are a road race and a media analysis:

Neither tells you the odds of finishing a hilly 50km trail race. Anyone quoting you a number is guessing.

How much should you train?

The studies agree on direction, not dose.

In 23,996 Two Oceans starters (Sewry and colleagues, Journal of Sports Medicine and Physical Fitness, 2022), not finishing was associated with fewer years of running, less training and racing per week, and a lower average weekly training distance. In 500 finishers and non-finishers of two 161km trail races (Hoffman and Fogard, International Journal of Sports Physiology and Performance, 2011), a faster finish time was associated with higher training volume. Both are self-reported and observational, so they show an association rather than proof that more kilometres cause a finish. Neither abstract gives a number of kilometres to aim for.

What the Western States 100 organisers publish is for a different race, but it is instructive. Their training page, written by John Medinger, sets out a sample week in hours rather than kilometres and says "many ultra runners can handle up to about six hours of running without significantly breaking down their muscles". It also says it is not necessary to run 161km a week: "Many runners are able to finish on about half this amount." It calls back-to-back long runs "a somewhat risky tactic" to be used "carefully and judiciously, if at all". That is one organiser's advice for a 161km race, not a study, so do not copy the numbers into a first 50km.

Here is the part we can support with a number. In a prospective cohort of 5,205 runners (Schuster Brandt Frandsen and colleagues, British Journal of Sports Medicine, 2025), a single session more than 10 percent longer than your longest run in the previous 30 days was linked to more overuse injuries: a hazard rate ratio of 1.64 for a 10 to 30 percent jump, 1.52 for 30 to 100 percent and 2.28 for more than 100 percent. Week-to-week jumps in distance showed no such link. The caveats: injuries were self-reported, and these were general runners, not ultra runners.

So: build your longest run gradually, one step at a time, rather than leaping from a 2 hour run to a 5 hour one. The old weekly "10 percent rule" is weaker than it sounds. A one-year study of 874 novice runners (Nielsen and colleagues, 2014) found no statistically significant difference in injury across weekly progression groups.

The nearest thing to a number comes from a coach, not a trial. Jason Koop, director of coaching at trainright.com and author of Training Essentials for Ultrarunning, told Suunto in 2016 that the returns from extra volume diminish "after about nine hours per week", and that for a 50km or 80km race about nine hours a week for six to eight weeks is enough. That is one coach's convention. It has not been tested in a study.

What does getting the build wrong look like? Trail Runner magazine asked three elite runners what they got wrong in their first 80km race, the 2017 North Face Endurance Challenge championship near San Francisco. Alicia Shay, a two-time NCAA 10km champion who finished 6th, said her longest run in the build was about 34km and never over three hours, "so I simply wasn't used to being on my feet for so long to have the confidence of how to pace myself." YiOu Wang, a 2:38 marathoner who dropped out at about 48km with a hamstring problem, said she had been overly ambitious to attempt her first 50km and her first 80km race in the same year: "I needed more time to recover." That is the hindsight of elite athletes after one race, not a study, but both mistakes were made in the build rather than on the day.

Do hills and vertical matter?

Almost certainly, but the evidence is mechanistic, not outcome-based. A review of graded running (Vernillo and colleagues, Sports Medicine, 2017) found that uphill running costs more energy than level running, while downhill running reduces energy cost until about a 20 percent downhill grade and then increases it, with heavier eccentric loading on the legs. In a 106km mountain ultra (36 athletes, Martinez-Navarro and colleagues, International Journal of Sports Medicine, 2026), uphill peak oxygen uptake alone predicted 79 percent of the variation in race time.

None of that tells you how much vertical to train or whether it changes your chance of finishing. It does suggest training on terrain that resembles the race.

Is strength training worth it?

A reasonable bet, with an asterisk. A meta-analysis of 5 controlled trials in 93 high-level middle and long-distance runners (Balsalobre-Fernandez and colleagues, 2016) found that strength and plyometric work 2 to 3 times a week for 8 to 12 weeks improved running economy with a large effect (standardised mean difference -1.42). A meta-analysis of 6 randomised trials (7,738 participants aged 12 to 40; Lauersen and colleagues, British Journal of Sports Medicine, 2018) found strength training cut sports injuries to roughly a third (relative risk 0.338), but across mixed sports and mostly younger athletes.

Neither studied ultra runners. If your calendar is full, this is the thing we would drop last.

How should you fuel?

Sources disagree, so be sceptical of anyone with one number.

Why the gap? A 2019 review (Costa and colleagues) put it plainly: whether 90g an hour is tolerable in an ultra is "questionable from a practical and gastrointestinal perspective".

The way through is gut training. A systematic review of 8 studies (Martinez and colleagues, Sports Medicine, 2023) found gut discomfort fell by an average of 47 percent after 2 weeks of repeated carbohydrate feeding in 2 of the studies, and in a 2-week randomised trial of 18 endurance runners (Miall and colleagues, Scandinavian Journal of Medicine and Science in Sports, 2018), the group that took 90g of carbohydrate an hour during a daily one-hour run reported fewer gut symptoms afterwards, while the placebo group did not. The studies are small and short, but a 2025 position statement from Sports Dietitians Australia and the Ultra Sports Science Foundation supports it too.

What to do: start with what your gut tolerates, practise eating on your long runs, and raise the amount over weeks. Your race-day plan should be one you have already used.

How much should you drink?

Drink to thirst, never drink so much that you gain weight, and learn your own sweat rate.

Here is why. At California 161km races (Hoffman and colleagues, Medicine and Science in Sports and Exercise, 2013), 15.1 percent of the 669 finishers whose blood sodium was measured had exercise-associated hyponatraemia (blood sodium below 135 mmol/L), and the authors concluded that excessive sodium supplements and drinking beyond thirst are ill advised. Overdrinking was not the whole story, though: only 23.8 percent of the runners with hyponatraemia were overhydrated (no weight loss, or a gain) at the finish. A 2017 review by the lead author of the 2015 international hyponatraemia consensus statement (Hew-Butler and colleagues, Frontiers in Medicine) concluded that drinking to thirst prevents it in temperate conditions in events under about 17 hours, and that fluid volume, not sodium intake, drives final blood sodium.

There is a counterpoint. A 2018 review (Kenefick, Sports Medicine), funded by the Gatorade Sports Science Institute, favours planned drinking in activities over 90 minutes, particularly in the heat, with sweat-rate testing to keep body mass loss under 2 percent. The Kenefick review adds a rule of thumb that suits both camps: never drink so much that you gain weight.

The Ultra-Trail Australia and Tarawera organisers also tell runners to drink to thirst and skip NSAIDs.

What about caffeine, night running and sleep?

Caffeine: an ISSN position stand (Guest and colleagues, 2021) found benefits consistently at 3 to 6mg per kilogram of body weight, with side effects common at 9mg per kilogram, and said it may help cognition under sleep deprivation. The stand covers exercise in general and is not specific to ultra-length racing.

Night running: if your race has a night in it, you need a headlamp. The 2027 UTA100 starts at 5am with a 28 hour limit, and its mandatory gear list requires a headlamp and a small backup light. In a survey of 1,154 finishers of a 165km and a 111km race (Kishi and colleagues, 2024), 77 percent slept during the race, short naps under 30 minutes were the most common, and 80 percent reported symptoms of sleep deprivation including falls and hallucinations.

How long should you taper?

Up to about three weeks, with the volume cut and the intensity kept. A meta-analysis of 27 studies (Bosquet and colleagues, 2007) found the most effective taper was 2 weeks, cutting training volume by 41 to 60 percent while keeping intensity and frequency. A 2023 meta-analysis of 14 endurance studies (Wang and colleagues) concluded a taper of up to 21 days at that same volume cut is effective. The Western States organisers suggest starting 2 to 3 weeks out. None of these tested ultras directly: the performance measures were competitions, time trials and time to exhaustion.

What gear is mandatory?

Every event sets its own list, so read yours. As an example, the 2027 UTA50 requires a long-sleeve thermal top, a waterproof and breathable jacket with hood, a beanie or balaclava, a charged phone with a working SIM, a whistle, a space blanket or bivvy sack, a compression bandage, 2 litres of water capacity, 2 food bars or portions per leg, a hi-vis vest, a headlamp and a dry sack. The UTA100 adds more. See the official mandatory gear page for the current list.

Where does the evidence run out?

Honestly, in the places you most want an answer:

The case for gradual builds, gut training, drinking to thirst and a short taper is stronger than the case for any specific plan. That is the honest summary.

Is a first ultra for you?

It is a long day, and our view is that a comfortable 10km is the minimum starting point, not the finish line of your preparation. If a 10km is your longest run today, build up through half-marathon distance on trails before you enter a 50km. That is our judgement rather than a study, and you will find coaches who disagree.

Inspired? Rehearse it on a long day with a crew

The big day on our Blue Mountains trip is the Six Foot Track out-and-back: 28 to 30km with more than 1,000m of climbing on the return, with a shorter 15 to 16km turnaround. It is not an ultra, but it is a long day on the right kind of terrain to test your fuelling plan. Dates are to be confirmed and the price is an estimate of about $1,650 per person. Booking is not open yet, so register your interest and we will contact you first.

Register interest in Run the Blue Mountains

Further reading

Sources

Every study below was checked against its published abstract (or full text where linked) on 4 October 2026.

Register interest in Run the Blue Mountains

Frequently asked questions

What counts as an ultra marathon?

A 2009 World Athletics article defines an ultramarathon as any distance longer than the traditional 42.195km marathon. Common first ultras are 50km events, such as the 51km UTA50 at Ultra-Trail Australia.

How many weeks should I train for a first ultra?

No source we could verify gives a number backed by evidence. Observational studies link more weekly training and more running experience with finishing and with faster times, but none gives a tested number of weeks or kilometres for a first 50km. One named coach, Jason Koop, told Suunto in 2016 that about nine hours a week for six to eight weeks is enough for a 50km. Treat that, and any other specific number, as a coach's convention rather than a finding.

How many carbohydrates should I eat per hour in an ultra?

Sources disagree. A 2014 review proposed up to about 90g an hour of multiple transportable carbohydrates for ultra-endurance events, while a 2019 International Society of Sports Nutrition position stand recommended 30 to 50g an hour of carbohydrate within 150 to 400 kcal an hour from varied foods. In a 2015 Dutch study, 75 percent of the 45 ultra runners tracked on race day ate under 60g an hour. Start where your gut copes and build with practice.

How long should I taper before an ultra?

Two meta-analyses support a taper of about two weeks (Bosquet, 2007) and up to 21 days (Wang, 2023), with training volume cut by 41 to 60 percent while intensity and frequency are kept. Neither tested ultras specifically. The Western States 100 organisers suggest starting a taper 2 to 3 weeks out.

Should I drink to thirst in an ultra?

A 2017 review by the lead author of the international hyponatraemia consensus found that drinking to thirst prevents exercise-associated hyponatraemia in temperate conditions for events under about 17 hours. A funded counterpoint favours a planned fluid strategy based on your sweat rate. A sensible middle path is to drink to thirst, never drink enough to gain weight, and learn your own sweat rate.