Athlete summary
How much carbohydrate you should take during a ride depends on how long you are out there and how hard the session is — not duration alone.
| Ride length (ACSM skeleton) | Default CHO | Notes |
|---|---|---|
| < ~45 min | 0 g/h | Mouth rinse at most |
| ~1–2 h | ~30 g/h | Single-source CHO is fine |
| ~2–2.5 h | ~60 g/h | Still usually single-source |
| > 2.5 h | up to 90 g/h | ACSM baseline for an untrained gut |
| > 2.5 h + gut trained and Wilson gate | 100–120 g/h | Optional elite band |
Intensity / role periodization
UCI Sports Nutrition Project guidance (Jeukendrup et al. 2026) periodizes on-bike CHO to workload / role / stage, not a single duration band. The engine keeps the ACSM 0 / 30 / 60 / 90 ladder as the only legal steps and nudges one rung:
| Demand | How we detect it | Effect on ACSM band |
|---|---|---|
| Hard / race | Race day, mountain-role hint, quality roles (VO2 / threshold / tempo / …), or IF ≥ 0.85 | Step up one rung (capped at 90) |
| Easy endurance | Endurance/recovery role with IF < 0.75 (unknown IF) | Step down one rung on mid/long bands (≥60) — e.g. easy 4 h → 60, not auto-90 |
| Baseline | Everything else / unknown context | ACSM duration unchanged |
Published illustrations (not a fitted model): Froome Giro 2018 ~57 g/h easy vs ~96 g/h mountain; Strobel 2022 Grand Tour 41–106 g/h tracking energy expenditure; WT practice 90–120 g/h on the hardest stages only.
Intensity never unlocks ≥100 g/h. Rates above 90 still need the gut-trained / prefs capability and a Wilson context gate.
Lab GLUexo personalisation
When a lab peak exogenous glucose oxidation is logged, a glucose-only 90 g/h target may be replaced by peak ÷ 0.80 (Podlogar/Wallis 2025; mean 65 ± 10 g/h). See Personalised in-ride glucose from lab GLUexo (Podlogar/Wallis 2025). Dual-source 100–120 g/h race rates are not capped. Missing lab data leaves this page’s bands unchanged. Oxidation is never invented from body mass.
No protein during the ride
Protein during exercise is not recommended (no extra muscle protein synthesis in working muscle; can slow gastric emptying / CHO delivery — Jeukendrup 2026). The in-ride plan is CHO + fluid + sodium only. Post-ride protein (0.3 g/kg) stays on the recovery block.
Gut-trained 100–120 g/h (Wilson 2025 gate)
Peloton practice ≠ single-day TT evidence. Wilson Sports Med 2025 found no clear single-day performance upgrade for ≥100 g/h vs ~90 g/h. King 2018/2019 often saw 90 match or beat 100–112 g/h; Podlogar 2022 raised exogenous oxidation at 120 without extra glycogen sparing. Norte 2026 fatigued-CP dose-response is modelled in the durability engine — see Fatigued Critical Power and W′ after prior work.
The 100–120 g/h band is never the default. The engine only recommends it when:
- the ride is already long enough for the ACSM 90 g/h duration band, and
- race nutrition prefs set a
target_carb_g_per_hourabove 90 or your reported / practiced tolerance sits above 90 g/h, and - at least one Wilson context gate holds:
- stage race (multi-day / GT recovery & daily CHO matching),
- high daily hours (≥ ~5 h, or an explicit flag),
- REDs-risk / problematic LEA energy matching,
- fatigued-CP goal (unlocks the band; Norte CP-preservation copy when set),
- explicit athlete pref (
target_carb_g_per_hour> 90).
A one-day race or fondo stays ≤90 g/h even if gut-trained, unless a gate above is set. Gut-training progressions and race-week dress rehearsals use the same race target so you practice the rate you intend to race at. When fatigued_cp_goal is set and the schedule lands ≥100 g/h, the fueling schedule surfaces a cho_cp_preservation note (Norte 2026: CHO attenuates fatigued-CP decline; W′ unprotected — still not a one-day default).
Carbohydrate type (glucose:fructose)
Above ~60 g/h you need multiple-transportable carbohydrate (glucose + fructose) so both intestinal transporters are used:
| Rate | Recommended glucose:fructose |
|---|---|
| ≤ 60 g/h | Single-source OK (no fixed ratio) |
| 61–90 g/h | ~2:1 |
| > 90 g/h | ~1:0.8–1:1 (closer mix for high oxidation) |
Literature
- Thomas, Erdman & Burke — ACSM/AND/DC position stand "Nutrition and Athletic Performance", Med Sci Sports Exerc 2016 (duration bands, ~90 g/h ceiling for general guidance).
- Jeukendrup — exogenous carbohydrate oxidation and multiple-transportable mixes, Sports Med 2014 (~60 g/h threshold; 2:1 guidance).
- Jeukendrup, Redegeld, Martins, Whitfield, Burke, Mujika, Dolan, Gonzalez — “UCI Sports Nutrition Project: Race Nutrition for Road Cycling”, Int J Sport Nutr Exerc Metab 2026;36(3):215–232 (doi 10.1123/ijsnem.2025-0239) — periodize on-bike CHO by intensity/role/stage; protein not during exercise; little/no performance evidence above 90 g/h.
- Wilson — “A Narrative Review of the High-Carbohydrate Fueling Revolution (≥ 100 g/h) in the Professional Peloton”, Sports Med 2025 (doi 10.1007/s40279-025-02372-6) — no clear single-day performance upgrade vs ~90 g/h; peloton practice ≠ single-day TT evidence.
- King et al., Physiol Rep 2018; Eur J Appl Physiol 2019 — 90 g/h often matched or beat 100–112.5 g/h on TT outcomes.
- Podlogar, Bokal, Cirnski, Wallis, Eur J Appl Physiol 2022 — 120 vs 90 g/h: higher exogenous oxidation, unaltered endogenous use.
- Podlogar, Cooper-Smith, Gonzalez, Wallis, Performance Nutrition 2025 — personalised glucose dose from peak GLUexo ÷ 0.80.
- Norte et al., Scand J Med Sci Sports 2026 — 120 g/h attenuates fatigued-CP drop (durability engine + fueling
cho_cp_preservationnote). - Dudley-Rode et al., Eur J Appl Physiol 2025 — CHO attenuates VT1 / transition power drift (~6% without vs ~3% with).
- Fordyce / Team Sky, 2018 — Froome Giro stages (≈57 vs ≈96 g/h) as illustration, not a formula.
- Strobel et al., 2022 — Grand Tour weighed intake 41–106 g/h tracking EE.
- Jeukendrup gut-trainability reviews + UCI / elite practice — accustomed athletes routinely target 100–120 g/h on long hard stages with closer glucose:fructose ratios (performance evidence for single-day ≥100: see Wilson 2025).