Sodium bicarbonate: the oldest legal ergogenic aid, and the stomach problem

Sodium bicarbonate has been studied as a performance aid since the 1930s, which makes it older than almost everything else on the shelf and considerably older than the marketing around it. It is baking soda. It costs almost nothing, it is legal, and both the IOC consensus statement on dietary supplements and the Australian Institute of Sport's ABCD framework put it in the small top group of supplements with sound evidence behind them.

It also has a reputation for making people spectacularly unwell, which is fair, and which has kept it out of a lot of kit bags that it might legitimately belong in. The gastrointestinal problem is real and it is largely solvable.

The bigger issue is that most cyclists who buy it have no use for it. The mechanism only pays off in a narrow band of effort durations, and if your target event is a five-hour road ride, you are buying a solution to a problem you do not have.

What it does, mechanically

Hard exercise produces hydrogen ions inside the muscle faster than they can be removed. Accumulating H⁺ interferes with several steps of contraction and energy supply, and it is one of several things that make a four-minute effort feel the way it does. Your muscle has its own internal buffers, but it also exports H⁺ and lactate into the blood, largely through monocarboxylate transporters.

That export depends on a gradient. Raise the buffering capacity of the blood outside the muscle, and the gradient steepens, and the muscle can shift more H⁺ out per second. Sodium bicarbonate does exactly that: it raises blood bicarbonate concentration and pH before you start, so the extracellular compartment has more room to absorb what the muscle sends it.

Note what this is not. It does not increase your aerobic capacity, it does not make you use fuel more efficiently, and it does not do anything inside the muscle cell. It is an extracellular buffer, and it only helps when the rate of H⁺ efflux is the thing standing in your way.

The dose is per kilogram, and the number matters

The dose that has been tested most is 0.3 g per kg of body mass. For a 70 kg rider that is 21 g — a little over four level teaspoons of baking soda. Some protocols use 0.2 g per kg and get a smaller but real rise in blood bicarbonate with fewer symptoms. Going above 0.3 g per kg reliably increases gastrointestinal complaints and does not reliably increase performance, which is a poor exchange rate.

Two things follow from the size of that dose. First, it is a lot of sodium: sodium is about 27% of sodium bicarbonate by mass, so 21 g of bicarbonate carries roughly 5.7 g of sodium. That is well above a normal day's intake in one sitting, and it comes with fluid retention and typically 0.5–1 kg of transient weight gain. If you have high blood pressure or any reason to be careful with sodium, this supplement is not for you. Note also that this is an entirely different thing from race-day electrolyte practice, which our piece on electrolytes, sodium and cramp covers.

Second, the timing is individual. Blood bicarbonate peaks somewhere between 60 and 180 minutes after ingestion, and where your own peak falls is fairly repeatable but not predictable from anything obvious. The tidy recommendation is to test it, find your time to peak, and plan backwards from the start of the effort. The pragmatic finding is that when researchers compared individualised timing against a fixed 180-minute pre-load, performance did not differ. So a fixed window of two to three hours is a defensible default.

The stomach problem, and what actually helps

Sodium bicarbonate meets stomach acid and produces carbon dioxide. That is the bloating. It is also a large osmotic load arriving in the small intestine, which pulls water in and can produce urgent, unwelcome results. In the study that quantified this properly, symptom scores after a 0.3 g per kg bolus were considerably higher than placebo, with substantial variation — some participants reported nothing, others reported enough to ruin a race.

Three modifications have evidence behind them. Taking the dose with a high-carbohydrate meal of around 1.5 g per kg reduced symptoms without reducing the bicarbonate rise. Splitting the total dose into smaller portions over a couple of hours lowers the peak load on the gut. And delayed-release capsules, which pass the stomach intact, produced the clearest reduction in symptoms of anything tested — this is the development that has made the supplement practically usable for people who previously could not keep it down.

Why it does almost nothing for a five-hour ride

This is the part that gets skipped, and it is the part most readers need.

On a long endurance ride you are not, for the vast majority of the time, generating H⁺ faster than you can clear it. That is more or less the definition of riding below threshold. Nothing is accumulating for extracellular bicarbonate to help with, so there is no mechanism by which a pre-ride dose could improve your average power over five hours. What limits you there is carbohydrate availability, heat, and the slow erosion of power over time — see durability and your power at hour four for what that erosion looks like on a file.

There is a partial exception worth naming honestly. If your five-hour ride contains a decisive four-minute climb at hour four, the effort that matters is in the useful window. But a dose taken before the start will have largely dissipated by then, and taking 21 g of sodium bicarbonate at hour three-and-a-half, on a stomach already managing 90 g of carbohydrate an hour, is a bad idea for reasons that should not need spelling out. If you want to read about what that stomach is already doing, how to eat on a five-hour ride is the relevant piece.

Who this is actually for

• Track and criterium riders, and anyone whose race is decided by repeated efforts of 30 seconds to a few minutes. The repeated-effort case is where bicarbonate looks best, because the buffer is still working on the fourth surge. Your first criterium describes the effort pattern.
• Hill-climb specialists, where the event is three to eight minutes of maximal effort and a 2% change in mean power is the difference between placings.
• Pursuiters and prologue riders, for the same reason.
• Not time triallists at 40 km, not gran fondo riders, and not anyone still in a base phase. If you want to know where your own efforts sit, Moveee's power profile shows which part of your power–duration curve you actually compete on, which is the first question to answer before buying anything.

Stacking it, and a word on expectations

Bicarbonate and caffeine work through unrelated mechanisms, and the meta-analytic evidence supports both independently for intense endurance efforts. Combining them is reasonable, though the combined effect is not guaranteed to be additive and the combined gastrointestinal risk certainly is. Get each one right on its own first; caffeine for cyclists covers the timing and the habituation question.

Beta-alanine raises carnosine, an intracellular buffer, and pairs conceptually with bicarbonate's extracellular buffering. Studies of the combination exist and lean positive, but the effect sizes are small and the confidence intervals are wide. Do not build a plan on it.