Fuelling research

How many carbs an hour for an ultra?

There's a real number in the literature, and it's worth knowing. It's also an average of other people, and in trials that measured them, at least four things that were not the carbohydrate moved the outcome. Both halves matter, so both are here.

Reviewed 26 August 2026. Every figure links the paper it came from.

The short answer, with its limits attached

Guidance for long efforts clusters around 60 grams of carbohydrate an hour, rising to about 90 for long events. Reaching 90 depends on mixing glucose and fructose rather than one source, because the two use different intestinal transporters and a single source saturates sooner (Jeukendrup 2014 review).

Recent work in trained runners has gone higher. One trial tested 120 grams an hour over a mountain marathon, in elite male runners who had rehearsed up to 90 grams an hour at least twice a week for the four weeks beforehand (Viribay 2020 trial). The rehearsal isn't a footnote. It's part of what was tested.

So the honest answer is a range, a population, and a condition: 60 to 90 grams an hour is where guidance sits, higher figures come from trained athletes who trained the gut first, and none of it was measured on you.

Who these numbers were measured on

This is the part usually left out, and it changes how far a figure reaches. The 120 grams trial was elite male mountain runners. Much of the transporter work was trained cyclists, in laboratories, on stationary bikes, at steady intensity, for two to three hours.

An audit of 937 studies on carbohydrate for exercise found about 11 per cent of the 11,202 participants were women, and about 79 per cent of the studies used male-only samples (Kuikman 2023). Not one study earned the top grade for controlling ovarian hormones, though 13 of 197 relevant studies reached an acceptable standard.

None of that makes the numbers wrong. It makes them narrower than the way they're usually quoted, and it means a woman running a twelve-hour trail race is further from the tested population than the figure suggests.

Four things that moved the outcome more than the number

Same carbohydrate, same kind of athlete, one thing changed. Each was measured.

  • The polyols in the days before

    At a 56km trail ultramarathon, the runners worst affected had eaten more polyols in the days beforehand. Total FODMAP intake showed nothing on its own.

    PubMed 38637153
  • What was eaten the night before

    One high-FODMAP day raised gut symptoms during the next day's run. Randomised double-blind crossover, 18 endurance runners, two hours at 35 degrees.

    PubMed 31652404
  • How anxious you were on the start line

    In a prospective field study at a 56km ultramarathon, pre-race anxiety tracked with how the gut went, most strongly in the final 24 hours.

    PubMed 34897557
  • The hour of the day

    The same three-hour run at 9pm moved through the gut more slowly than at 9am, with more symptoms and a much larger cortisol rise. Crossover trial in 16 runners.

    PubMed 33065594

Four variables, one gel, and not one of them appears on anybody's Strava. It's why a number alone is a weak answer, and why the same runner can take the same product on two days and get two outcomes.

What the research does not say

It doesn't say which figure is yours. Every number above is a group mean, drawn from a sample you're probably not in, and studies that measured gut outcomes report how a group responded, not how to predict one person.

It also doesn't say a bad day was caused by the carbohydrate. Nothing that reads a finished run can establish that, including Absorb. What can be established is what a bad day coincided with, and whether that combination has come up before.

So the useful move isn't finding the right number and adopting it. It's knowing which conditions you've already run in with fuel on board, and which you haven't, before a start line asks.

Where Absorb comes into it

Absorb is a fuelling memory for endurance and ultra runners. It reads the runs already on your Strava, records what happened on each one, one tap mid-run or two after, keeps the conditions each run happened in, and shows which of them you have notes for and which you don't. It never sets you a number, and it doesn't claim to know why anything happened.

See how the weekly loop works Questions runners ask

Population averages from published literature, not individual nutrition advice. Absorb is not a medical or dietary service.

Join the list

Be there when it opens

Absorb isn't open yet. Leave your email and we'll tell you on the day it is.

Two emails, ever: the day it opens, and one about the founding places.

For people aged 16 or over.