Everyone knows you can load baking-soda-style buffers before a hard effort. Almost nobody has tested what happens when you take them during a long ride. A new human crossover trial in trained cyclists did exactly that, and the late-ride result was hard to ignore: +12 W in a 12-minute time trial after three hours of riding. The catch is that the sprint and 2-minute results didn't budge, and the sample was tiny.
This is a human study in 11 trained cyclists, not an animal or lab-cell experiment. Here is what was actually found.
Why This Study Is a First
Sodium bicarbonate (SB) has a track record for high-intensity efforts lasting roughly 1 to 60 minutes. Until now, the research has centered on taking it beforehand. According to the authors, no earlier study had tested SB during exercise, or whether it could blunt the fatigue-related decline in the power-duration curve over a long effort. That decline is what coaches increasingly call durability: how much of your power you still have after hours of work.
The study, published in Scandinavian Journal of Medicine & Science in Sports (Norte et al., 2026), tested bicarbonate and carbohydrate together in a novel gel format.
How the Trial Worked
Eleven trained cyclists completed a randomized, crossover design. The steps were:
- Baseline visit: a non-fatigued power profile made up of three 6-second sprints, a 2-minute time trial (TT) and a 12-minute TT.
- Two experimental trials: 3 hours of cycling at 95% of gas exchange threshold, with carbohydrate at 80 g/h.
- Two conditions: carbohydrate alone (CHO), or carbohydrate plus bicarbonate (CHO + BI), using 18 g sodium bicarbonate and 2 g potassium bicarbonate in total.
- The test: the same power profile repeated after the 3-hour ride, so riders were tested in a fatigued state.
The Headline Numbers
12-minute TT: Power was 12 W higher with CHO + BI than with CHO alone (p = 0.01, Hedges' g = 1.07). That is a large effect size, and the only statistically significant performance result in the paper.
Everything else was a shrug:
- 6-second sprint power: +12 W (p = 1.00, g = 0.18), not significant
- 2-minute TT: +6 W (p = 0.888, g = 0.32), not significant
The sprint difference matches the 12-minute difference in raw watts, but it came with a tiny effect size and very high variability relative to the signal, so it can't be treated as a real gain. The benefit shows up in the sustained, threshold-ish effort, not in short, explosive work.
What About Your Stomach?
GI distress is the usual reason people abandon bicarbonate. Here, there were no differences in gastrointestinal symptoms between conditions. The authors describe the discomfort as minimal. Delivering the dose inside a carbohydrate gel, spread across the ride, appears to have been tolerable in this group.
Why It Matters for Long-Course Racing
If you race triathlon, gran fondos or long road events, the decisive moments often come late, when you're depleted and need to push. This study suggests that buffering support during the ride, not just before it, might help preserve sustained power at that point. The authors frame it as potentially attenuating fatigue-related decline, though a single 12-minute test can't prove a durability mechanism by itself.
Practical Takeaways
Dose used in the study: 18 g sodium bicarbonate plus 2 g potassium bicarbonate across the 3-hour ride, alongside 80 g/h of carbohydrate in gel form. If spread evenly, that works out to about 6 g of SB per hour (our arithmetic, not a figure from the paper). It also delivers roughly 5 g of sodium from the SB alone.
Timing: Ingested during exercise, as part of the fueling plan, rather than as a pre-race bolus. The abstract doesn't give the exact feeding schedule, so don't assume intervals beyond "throughout the 3 hours."
Who it might suit:
- Trained cyclists, triathletes and long-course racers who already fuel at high carbohydrate rates
- Athletes whose events end with a sustained 10 to 15 minute hard effort
- People willing to trial it thoroughly in training first
Who should skip it or be cautious:
- Anyone told to restrict sodium, or with blood pressure, kidney or heart concerns. Talk to a clinician first (general caution, not from the paper).
- Riders with a history of GI trouble who haven't tested a bicarbonate protocol in training
- Sprinters and short-event specialists. The sprint and 2-minute results showed no meaningful benefit here.
- Anyone expecting a magic bullet. This is one small trial.
Limitations You Should Know
- Small sample: only 11 trained cyclists, so the findings need replication in larger and more varied groups, including women and different performance levels.
- Narrow benefit: gains appeared in the 12-minute TT only. Sprints and the 2-minute TT were not significantly improved.
- Single protocol: a 3-hour ride at 95% of gas exchange threshold with 80 g/h carbohydrate may not mirror your race.
- Product provision: the nutrition products were supplied by MNSTRY, and Siemens Healthineers provided the blood analyzer. The authors state neither company had a role in the design, data collection, analysis, interpretation or the decision to publish. One author is affiliated with Precision Fuel & Hydration.
- Lab, not race: time trials after a controlled ride are not the chaos of a real event.
The Bottom Line
This is the first look at bicarbonate taken during prolonged cycling, and it found a meaningful 12 W late-ride boost in a 12-minute effort without extra GI symptoms. It is promising, but small and narrow. Treat it as a strong reason for more research, and a reasonable experiment for endurance athletes to discuss with a qualified professional, not as a guaranteed upgrade.
