Carb Loading for Runners: The Science of Pre-Race Fueling

The original carb loading protocol was developed in the 1960s by Swedish physiologist Per-Olof Åstrand and colleagues. It involved a week of structured pre-race nutrition: three days of glycogen-depleting exercise and low carb intake, followed by three days of high carb intake and rest. The protocol roughly doubled muscle glycogen stores compared to baseline. Runners performed dramatically better in long events. The technique was a genuine breakthrough.

It was also miserable. The depletion phase made runners exhausted, irritable, and prone to colds during the taper they were supposed to be benefiting from. Most amateurs who attempted it described the week before their marathon as the worst of training, not the best.

Modern research has shown that the depletion phase is unnecessary. A simple 24–36 hour high-carb intake before a long race elevates muscle glycogen stores to nearly the same level as the classical protocol, without any of the depletion-phase downsides. This is the modern carb loading protocol, and it actually works — when the carb numbers are real.

This guide covers what carb loading does physiologically, who actually needs it, the modern protocol with specific gram targets, food choices that work, and the common mistakes that turn a productive pre-race day into a digestive disaster.

What carb loading actually does

The body stores carbohydrate as glycogen in two places: muscle (roughly 80% of total stores) and liver (roughly 20%). Total typical glycogen stores in an untrained adult: 300–400 grams, equivalent to about 1,200–1,600 calories. In trained endurance athletes, glycogen storage capacity is higher — typically 400–600 grams.

During hard running, the body burns a mixture of carbohydrate (glycogen and blood glucose) and fat. The proportion shifts based on intensity: at easy effort, fat dominates; at race intensity, glycogen contributes 60–80% of energy. The body cannot run at race intensity on fat alone — there is a hard ceiling on the rate at which fat can be oxidised.

When muscle glycogen runs low, the body has to slow down. The phenomenon distance runners call "the wall" or "bonking" is largely a glycogen depletion event. For a runner who has not deliberately maximised pre-race glycogen, the wall often arrives between 30 and 35 km of a marathon — exactly where the back-half slowdown becomes infamous.

Carb loading raises starting glycogen stores by approximately 25–50%. The practical effect: the wall is delayed (or in some cases avoided entirely), pace decay in the back half of the race is reduced, and average finish time improves measurably. A 2011 review in Sports Medicine (Burke et al.) summarised the research: carb loading produces an approximately 2–3% performance improvement in events lasting 90 minutes or longer.

Who needs to carb load and who doesn't

The 90-minute rule is the standard cutoff. Events shorter than 90 minutes typically don't deplete glycogen stores to performance-limiting levels in well-fed runners. Events longer than 90 minutes increasingly benefit from elevated starting stores.

• 5K, 10K (under 50 minutes for most amateurs): Modest pre-race carb intake is sufficient. A regular meal the night before and a normal breakfast is enough.
• Half marathon: Borderline. Faster runners (under 90 minutes) may not need aggressive loading. Slower runners (over 2 hours) benefit from moderate loading — 8 g/kg the day before is reasonable.
• Marathon: The classic event where carb loading delivers its largest benefit. Full 24–36 hour protocol at 10–12 g/kg per day.
• Ultramarathon: Carb loading helps but is supplemented by in-race fueling. The longer the event, the more in-race nutrition matters relative to pre-race stores.

For races below the 90-minute threshold, attempting an aggressive carb load often produces digestive discomfort without performance benefit. The body simply doesn't need that much pre-race fuel for shorter events.

The modern protocol: the actual numbers

Current sports nutrition guidelines (International Society of Sports Nutrition; the Burke et al. position statements) recommend the following for endurance events of 90+ minutes:

These are large numbers — significantly higher than typical daily carb intake even for endurance athletes during normal training. Most runners attempting their first deliberate carb load are surprised by how much eating it actually requires.

To translate: 700 grams of carbohydrate is roughly equivalent to:

• 4 cups of cooked rice (200 g)
• 3 large bagels (180 g)
• 2 large bananas (60 g)
• 4 cups of cooked pasta (240 g)
• 2 cups of fruit juice (60 g)

That's roughly 740 g — and represents a substantial eating day. Most runners who claim to "carb load" before a marathon are actually eating 4–6 g/kg per day — half the actual target. The result is partial glycogen elevation and a smaller-than-possible race-day benefit.

For the broader nutrition context, see running nutrition guide.

What to actually eat

The carb sources that work for loading are those that are easy to consume in large quantities without producing digestive issues. The major considerations:

Low fibre during the final 24 hours. Whole grains, beans, broccoli, large salads — all useful in normal training nutrition — slow digestion and increase pre-race gut discomfort. The final pre-race day is the wrong time to eat large amounts of fibre. Switch to refined carbs: white rice, white pasta, white bread, regular potatoes (peeled), bananas, white sourdough.

Moderate protein. Some protein with each meal (40–50 g total per day) helps maintain satiety and provides amino acids. But protein-heavy meals (large steaks, heavy meat dishes) displace carbs and slow gastric emptying. Lean protein like chicken or fish, in modest portions, is the standard.

Lower fat than usual. Fat slows digestion and competes for stomach capacity. Heavy fat meals (creamy pastas, fried foods, lots of cheese) feel filling and reduce how much carb the runner can actually consume.

Liquid carbs help. Fruit juices, sports drinks, and sugar-sweetened beverages contribute carb without filling the stomach. A glass of fruit juice with lunch and dinner adds 50–80 g of carb each.

Test in training. Any food that has not been tested during long training runs is a race-day risk. Whatever the runner ate before successful long runs in training is a much safer choice than a "carb-loading dinner" they have never eaten before.

Race morning

The morning of a long race introduces a different challenge: providing additional fuel without leaving food in the stomach during running. The standard guidance:

• Wake 3–4 hours before race start. Earlier than feels natural, but provides time to digest a meal.
• Pre-race breakfast: 1–4 g carb per kg body weight. For a 70 kg runner: 70–280 g of carbohydrate. The exact amount depends on race-morning timing and individual digestion tolerance.
• Familiar foods only. The morning of a goal race is not the time to try a new breakfast.
• Liquid options for runners with race-day nerves. Many runners struggle to eat solid food before races due to anxiety. Sports drinks, gels, or carb-rich smoothies work as alternatives.

A standard race-morning breakfast example for a 70 kg runner: 2 cups oatmeal with honey and banana (90 g carb), a slice of toast with jam (40 g), a glass of sports drink (30 g). Total: 160 g — within the 1–4 g/kg target. Eaten 3 hours before the gun, this food is largely digested before running starts.

For the final 30–60 minutes before the race, small amounts of easily-digested carbohydrate — a gel, half a banana, sports drink — can top off blood glucose without sitting heavy. Avoid significant solid food in the final hour.

Water and electrolytes during loading

Glycogen storage requires water — roughly 3 grams of water per gram of glycogen stored. A successful carb load increases body water by 1–2 kg, which is part of where the pre-race "carb load weight gain" comes from.

Practical implications:

• Water intake naturally rises during a proper carb load — the body is thirsty more often.
• Don't try to "stay light" by restricting water during loading. The water gain is part of the productive effect.
• Sodium intake should be normal to slightly elevated during loading. Salty foods at meals, electrolyte drinks alongside meals, and salt on rice/pasta all help retain the extra water.
• Urine that becomes pale yellow is the target. Dark urine during loading indicates insufficient fluid.

The weight gain question

A typical successful 36-hour carb load produces 1–2 kg of weight gain. Almost all of this is water bound to glycogen. The runner is not heavier in any race-relevant sense; the additional mass is functional fuel.

This trips up runners who have been weight-conscious during training. Stepping on the scale 24 hours before the race and seeing a number 1.5 kg higher than the day before can be psychologically difficult. The number means the loading is working. The water leaves the body during the race as glycogen is consumed.

Some runners deliberately avoid the scale in the final 48 hours pre-race for this reason. A reasonable practice.

Common mistakes

Eating more than usual rather than specifically more carbs. The classic "carb loading dinner" of a giant bowl of pasta with rich meatball sauce produces a big stomach without a big glycogen load. The protein and fat in the sauce compete for stomach capacity. Lean protein in modest portions, lots of carbs, low fat is the productive structure.

Loading on the wrong day. Some runners attempt to load only the day of the race. Glycogen storage from a single meal is slow; the elevated stores from carb loading require 24–36 hours of sustained intake. Day-of loading produces a full stomach without elevated muscle glycogen.

Loading without testing. Race day is not the time to discover that this particular brand of sports drink doesn't sit well, that pasta sauce produces reflux, or that white bread produces bloating. Practice the loading protocol — including specific food choices — before a long training run, ideally twice, in the weeks before the race.

Continuing high training during loading. The taper exists for a reason. Hard training the day before a race depletes glycogen as fast as the loading replaces it. The taper and the carb load are designed to work together; doing one without the other halves the benefit.

Adding new foods. A runner whose normal diet doesn't include a particular high-carb staple should not introduce it during the pre-race load. Gut tolerance is individual. The day before a marathon is not the day to discover personal intolerance to a new bread or pasta.

Drastically over-eating. The target is high carb intake, not a binge. Stomach distension produces discomfort that lingers into race morning. Multiple medium meals every 2–3 hours produce better outcomes than two enormous meals.

Skipping the load entirely because "I had pasta last night." One pasta meal is not a carb load. The protocol is 36 hours of sustained high carb intake, not a single themed dinner.

In-race fueling: the other half

Even maximum pre-race glycogen stores are not enough for the full marathon. Most runners can hold roughly 90–120 minutes of race-pace running before glycogen drops to performance-limiting levels. After that, in-race carbohydrate intake bridges the gap.

The standard guideline for events longer than 90 minutes:

• 30–60 grams of carbohydrate per hour for events 1–2.5 hours
• 60–90 grams per hour for events 2.5+ hours, in a 2:1 glucose-to-fructose ratio for better absorption

This translates to roughly 1–2 gels per hour, or equivalent in chews, sports drinks, or whole foods. Some runners struggle to consume the higher end of the range; gut training during long training runs (deliberately practicing the in-race fueling protocol) meaningfully improves tolerance.

In-race nutrition is a topic in itself. The key point for carb loading is that pre-race loading and in-race fueling are complementary, not interchangeable. Maximum pre-race stores still need to be topped up during a marathon; modest pre-race stores cannot be fully compensated for by in-race fueling alone.

Special cases

Low-carb / keto-adapted runners. Runners following long-term low-carb diets have shifted toward higher fat oxidation rates, but their maximum race-pace performance is still limited by carbohydrate availability. The research on fully fat-adapted runners shows that reintroducing carbohydrate before a race produces noticeable performance gains even in adapted athletes. For race-focused low-carb runners, a 48-hour carb reintroduction before goal races is reasonable.

Runners with gut sensitivity. Some runners cannot tolerate the 10–12 g/kg target — gut volume is the binding constraint, not appetite. For these runners, a lower target (6–8 g/kg) over a longer window (48–72 hours) produces partial benefit without the digestive backlash. Liquid carb sources (sports drinks, juice, maltodextrin-based recovery drinks) help meet the target with less stomach distension.

Hot-weather races. In heat, the body uses more glycogen at any given pace (heat stress is metabolically expensive). Pre-race carb loading is, if anything, more important in hot conditions. Hydration alongside loading also matters more. See running in the heat.

The 36-hour countdown

The bigger picture

Carb loading is the single most evidence-supported nutrition intervention in distance running. Done well, it produces a 2–3% performance improvement — equivalent to weeks of additional training. Done poorly (or skipped entirely), it leaves significant performance on the table.

The protocol is straightforward but unforgiving. The 10–12 g/kg target requires deliberate eating, not just "having pasta the night before." The 24–36 hour window matters. The food choices matter. The willingness to practice before race day matters.

For runners targeting their first marathon, getting the carb load right may have a larger effect on finish time than any single training session in the final two months. For runners targeting a PB, it is one of the few remaining variables under direct control on race week.

For broader race-week preparation, see how to taper for race day. To lock in your goal pace and per-km splits before race day, use the marathon pace calculator.