Beta-Alanine: Real, Small, and Only for the Two-Minute Effort. Twenty Trials Say It Does Nothing to Your Body.

This is the hundredth claim we have checked on this site. It is also one of the few where the answer is yes.

Beta-alanine works. Two independent meta-analyses, a decade apart, agree that it does something measurable. There is a plausible mechanism, a serious safety review behind it, and no sign of anybody having to squint at the data to find the effect.

So this page is not about whether it works. It is about the four qualifiers that come attached, because every one of them changes whether you should bother.

What it does, and how much

The definitive analysis pooled 40 studies covering 1,461 people, double-blind and placebo-controlled throughout. [1]

Overall effect size: 0.18. On the scale researchers use, 0.2 is considered small, 0.5 moderate and 0.8 large — so 0.18 is at the bottom end of small. Real, and real is not the same as big.

The earlier meta-analysis put a friendlier number on it: a median improvement of 2.85% in whatever was being measured. [2] That is the honest headline. Not a transformation. A nudge that a competitive athlete would take and most people would never notice.

The distinction almost nobody explains

Here is the part worth the price of admission, because both meta-analyses found it and it rarely survives into a product description.

Researchers separate two things that sound identical:

Exercise capacity is how long you can keep going before you stop. Ride at a fixed hard pace until you cannot any more, and see what the clock says.

Exercise performance is how fast you complete a set task. Ride four kilometres as quickly as you can, and see what the clock says.

Beta-alanine improves the first. The second does not reach significance. In the big analysis, capacity came out at an effect size of about 0.50 — a genuinely moderate effect — while performance came out at 0.11 with a range that comfortably included zero. [1] The earlier analysis found exactly the same split. [2]

Which means beta-alanine may help you hang on longer at a given intensity, without necessarily making you faster over a fixed distance. Those are different things to buy, and only one of them has been demonstrated.

And only in a particular window

Efforts lasting one to four minutes improved most clearly. Longer than four minutes improved more weakly. And efforts under 60 seconds showed nothing at all. [2]

That fits the proposed mechanism, which is a point in its favour. Beta-alanine raises muscle carnosine, and carnosine helps mop up the acidity that builds during hard efforts. A five-second lift never gets acidic enough for that to matter. A four-minute effort absolutely does.

So: rowing, 400 to 1500 metre running, hard intervals, the back half of a tough set. Not a one-rep max, and not a long slow run.

Two more things the label will not tell you

More does not work better. The big analysis specifically tested whether the total amount ingested moderated the effect. It did not. [1] Nor did training status — trained and untrained people responded similarly.

It does nothing to your body. A separate meta-analysis pooled 20 trials and 492 people looking at body composition. Body mass, fat mass, body fat percentage, fat-free mass — every single one came back null, with zero heterogeneity between studies, which is about as clean a null as this field produces. Slicing by exercise type, duration and dose changed nothing. [3]

If it is in your pre-workout and you are taking it to look different, that part is not happening.

The tingling, and whether it matters

If you have taken a pre-workout you have felt it: the prickling across the face, neck and hands about twenty minutes in. That is beta-alanine, and it has a name — paraesthesia.

A systematic risk assessment pooled 101 human studies and 50 animal studies looking specifically for harm. [4] The finding is unusually clean: paraesthesia was the only side effect reported, roughly nine times more likely than on placebo. People taking it dropped out of trials at the same rate as people taking placebo, so it was not driving anyone away.

There was a small rise in one liver enzyme, with the reviewers noting the averages stayed well within normal clinical ranges. No effect on muscle taurine, which had been a theoretical worry. Their conclusion: at the doses studied, it does not adversely affect the people taking it.

So the tingle is harmless and it is also the most reliable thing about the product. Worth knowing that feeling it tells you the dose arrived. It does not tell you it is working.

One gap worth flagging

Most of this evidence comes from men, or from mixed groups. A 2026 review looked specifically at women: 11 trials, 312 participants. Time to exhaustion did improve, at a moderate effect size. But every other outcome was imprecise, and the reviewers graded the certainty of all outcomes as very low after downgrading for risk of bias, imprecision and publication bias. [5]

That is not evidence it works differently in women. It is evidence that nobody has looked hard enough to say.

The verdict

Supported. Beta-alanine does something real, replicated by independent teams a decade apart, with a mechanism that predicts exactly the pattern the data shows and a safety review covering a hundred human trials.

It is Supported rather than Established because the overall effect is at the small end of small, the benefit is confined to one type of measurement inside a narrow duration window, and the evidence in women is graded very low certainty.

Worth it if you compete or train hard in efforts lasting roughly one to four minutes and a couple of percent matters to you.

Not worth it if you are lifting heavy for low reps, running long and slow, or hoping it changes how you look. Twenty trials say the last one is not happening.

Related: creatine, the other supplement in this aisle that survives contact with the evidence — along with whether it is safe and how much you actually need. The full scoreboard is on supplements for muscle growth, ranked by evidence.

If you want to know what earns each rating, we wrote down how we read a study, and everything we have checked — all hundred of them — is in one place.