ArticleπŸ“… 07.05.2026⏱ 10 min readπŸ€– AI Research

Carb Loading for Marathon Runners: Science-Backed Protocol and How Many Grams You Need

Carb loading is the only dietary strategy in endurance sport whose effectiveness has been confirmed in controlled trials for over 50 years. Done right, it boosts muscle glycogen stores by 50–100% and pushes back the wall β€” the moment of depletion at kilometer 30–35. Done wrong, it triggers diarrhea, adds 1–2 kg of dead weight, and undoes weeks of training.

What carb loading is and why it matters

Glycogen is the storage form of glucose in muscles and the liver. While running at marathon intensity (75–85% of VO2max), the body burns mostly carbohydrates. Stored glycogen lasts about 90–120 minutes of hard work. A runner aiming for a 3:30–4:00 finish inevitably hits the point when glycogen runs out and pace collapses β€” that is the marathon wall.

The goal of carb loading is to reach the start line with glycogen stores fully topped off. According to the American College of Sports Medicine (ACSM), an untrained person stores roughly 80–100 mmol/kg of dry muscle mass. After a proper load, this number climbs to 180–200 mmol/kg.

How many carbs you need: precise targets

The IOC Sport Nutrition Institute and ACSM publish aligned recommendations. They depend on duration of effort and athlete weight.

Race durationCarb targetTime frame
Up to 60 min (10K)5–7 g/kg/dayregular nutrition
1–3 hours (half marathon)7–10 g/kg/day2 days before
Over 90 min up to 4 hours (marathon)10–12 g/kg/day1–3 days before
Ultra (over 4 hours)10–12 g/kg + on-course carbs2–3 days before

Take a runner weighing 70 kg preparing for a marathon. Target: 10 Γ— 70 = 700 g carbs per day. That is a huge number β€” about 2,800 calories from carbs alone, which for most people exceeds their normal intake. So fats are deliberately cut to 0.8–1 g/kg, while protein stays at 1.4–1.6 g/kg.

πŸ’‘ Important: most amateurs overestimate how many carbs they actually eat. A plate of pasta is 60–80 g of carbs, not 200. Track your food for 2–3 days before loading to know your real starting point.

Classic protocols: which one works best

Bergstrom protocol (1967) β€” outdated

Swedish physiologist Jonas Bergstrom proposed a depletion-based scheme: a week before the race, an exhausting workout drains glycogen, then 3 days of low-carb intake (under 100 g/day), then 3 days of loading. The idea was supercompensation. In practice, the depletion phase weakens immunity, raises injury risk, and is psychologically brutal. Modern Cochrane and ACSM reviews do not recommend the classic version for amateurs.

Sherman-Costill modified protocol (1981) β€” gold standard

No depletion phase. Gradual carb increase against a backdrop of declining training volume (taper).

Day before raceCarbs (g/kg)Training
βˆ’65–6long run 60–90 min
βˆ’55–6easy 40 min
βˆ’45–6rest or 30 min
βˆ’38easy 30 min
βˆ’210rest
βˆ’110–12rest or 15-min shakeout
Race day2–4 g/kg, 3–4 hours pre-racemarathon

One-day Fairchild protocol (2002)

The day before the race: 10–12 g/kg of carbs plus a 3-minute all-out sprint that morning. The study showed a 90% glycogen rise in 24 hours. Suitable for experienced runners with good weight control. For beginners, the risk of error outweighs the benefit.

Which foods to choose

Main rule: at the loading stage you need easily digestible carbs with minimal fiber and fat. Whole grain bread and quinoa are excellent food in general, but 700 g of carbs from them means 50 g of fiber and a guaranteed bathroom collapse on race morning.

FoodCarbs per 100 gFiberSuitable for loading
White rice, cooked28 g0.4 gideal
Wheat pasta, cooked30 g2 gideal
Boiled potato17 g1.3 ggood
White bread49 g2.5 ggood
Banana23 g2.6 gexcellent
Honey82 g0 gexcellent (add-on)
Brown rice23 g1.8 gnot the best choice
Whole oats, cooked12 g1.7 gonly at βˆ’3 days
Legumes18–22 g7–8 gavoid on βˆ’2/βˆ’1 days

Food composition data: USDA FoodData Central.

Sample menu for the day before a marathon (70 kg runner, 700 g carb target)

Total: ~700 g carbs, ~95 g protein, ~30 g fat, around 3,500 calories. It sounds like a lot, but that is exactly what muscles need. Eat often and in small portions; do not try to fit this into three big meals.

Race morning

Breakfast 3–4 hours before the gun: 1–4 g of carbs per kg of body weight. For a 70 kg runner that is 70–280 g. Classics: oatmeal in water, white bread with honey, banana, a strong cup of coffee (3–6 mg caffeine/kg β€” scientifically validated for marathon performance per ACSM).

15–30 minutes before the start β€” 30 g of fast carbs: a gel, sports drink, or chews. No new foods: anything you plan to consume on race day must be tested on a long run 4–6 weeks before the race.

⚠ Never try a new gel, new pasta brand, new supplement, or exotic cuisine the night before a marathon. GI distress is the cause of DNF for 30–50% of amateur runners according to finisher surveys. It is the most common β€” and most preventable β€” reason to drop out.

Myth busting

Myth 1: A plate of pasta for dinner = I'm loaded

One serving of pasta is 60–90 g of carbs. That falls short of the 700 g target by 8–10 times. Cramming a loading day into a single meal is physiologically impossible: stomach capacity is limited. Loading is an all-day, every-2-hours pattern, ideally lasting 24–36 hours.

Myth 2: I'll gain fat from this much carbs

Each gram of stored glycogen pulls 2.7–3 grams of water with it. With a full load, weight may rise by 1.5–2 kg β€” that is water and glycogen, not fat. After the race, that water leaves within 24–48 hours. This is normal and desired: an extra 1.5 liters of water in the muscles helps fight dehydration on the course.

Myth 3: A keto marathoner does not need to load

A meta-analysis by Burke and colleagues (2017, Journal of Physiology) showed that on a long-term keto diet mitochondria adapt to fat oxidation, but running economy at marathon intensity drops by 5–8%. That means 10–15 minutes lost on a marathon. For competitive results, a ketogenic approach loses to traditional carb loading.

πŸ’‘ Remember: carb loading only works for races longer than 90 minutes. For 5K or 10K it is pointless β€” your normal glycogen reserves will not be depleted anyway.

Common mistakes

What science says: key sources

Modern recommendations rest on consensus documents:

Bottom line: the loading effect is well documented, but execution requires precision. The biggest failures are undershooting the carb target, eating too much fiber, and experimenting with new foods on race weekend.

Track loading carbs without spreadsheets

Snap a photo of your plate and an AI will name the carbs, protein, fat, and calories. Especially handy on loading days, when you need to hit 700+ g of carbs without losing track.

Open @botnutraibot β†’