Citrulline malate is a permanent resident of pre-workout blends. The sales pitch is simple and sounds convincing: citrulline raises arginine availability, arginine feeds nitric oxide synthesis, nitric oxide dilates the vessels, more blood reaches the muscle. Add malate as a participant in the Krebs cycle. A pretty chain in which every individual link is true.

The trouble is that true links do not guarantee a result at the finish. In June 2026 a meta-analysis appeared that collected everything there is on this supplement — and the conclusion came out noticeably more modest than the advertising.

What was done

The paper was published in Nutrients. It is a systematic review with a three-level meta-analysis — a method that can correctly handle the situation where a single study yields several outcomes at once.

Scope: 30 randomised controlled trials, from which 138 effect sizes were extracted across 644 participants. Effects were pooled as Hedges' g in three-level random-effects models. Acute intake (a single dose before exercise) and chronic intake (a multi-week course) were analysed separately.

This is sound methodology. And here is the thing worth grasping: the more careful the method, the more honestly it exposes how weak the underlying data are.

What came out

The authors' final wording goes like this: the supplement may be associated with small, context-dependent improvements in performance, but the evidence base remains limited and uncertain.

The details:

Acute intake looks more consistent than chronic. A single dose before work has more going for it than a course.
In subgroup analysis among the acute protocols, a benefit emerged for aerobic endurance and short anaerobic tasks. But — and this is the key part — those findings did not survive the sensitivity analyses. Put plainly: nudge the calculation conditions slightly and the effect dissolves.
No significant differences between subgroups were found, not by sex, not by training status, not by dose, not by timing of intake. That is separately interesting: the stock advice — «take it 60 minutes before», «it only works in the untrained» — is not supported by the data.
Certainty on the GRADE system: low to very low.

That last point deserves translating from science into human. GRADE rates how much we can trust the effect estimate itself. «Very low» means the true effect is likely to differ substantially from the number obtained. That is not «the supplement doesn't work» — it is «we don't know».

And a separate study that stands beside it

In the same year, 2026, a 10-day randomised crossover trial in 20 moderately active young adults appeared: citrulline did not improve time to exhaustion during high-intensity cycling.

One null result proves nothing on its own. But it illustrates nicely what that «low certainty» is made of: the base consists of small studies, some of which show an effect and some of which do not.

How to think about this

There are two easy extremes here, and both are wrong.

Extreme one: «the meta-analysis showed a benefit, so it works». No — the benefit shown did not survive a robustness check, and the certainty was rated low.

Extreme two: «science has refuted it, throw it out». Also no. There was no refutation. There was an honest «the data are not sufficient for a confident conclusion».

The practical bearings that follow:

If you already take it and you like it — there is no reason to stop. Citrulline's safety profile is good, and the effect, if it exists at all, is small and is hardly getting in your way.
If you are choosing where to invest — there are supplements with an incomparably firmer base. Caffeine, creatine, beetroot nitrates and sodium bicarbonate are covered separately here, and each has an order of magnitude more evidence behind it.
Do not expect a noticeable gain. A «small, context-dependent» effect is not something you feel in a session. If you do feel something, it is most likely the caffeine from the same pre-workout blend.
Forget the fine points of the protocol. Since no differences by dose or timing were found, arguing about «40 minutes before or 60» is pointless.

The broader thought this was worth unpacking for: a plausible mechanism is not evidence. The chain «citrulline → arginine → nitric oxide → vasodilation» is flawless on paper, and that is exactly why it sells so well. What gets tested in the end is not the mechanism but the result over the distance.

Key points

The 2026 meta-analysis: 30 RCTs, 138 effect sizes, 644 participants, three-level model.
Conclusion: the effect is small and context-dependent, the evidence base limited.
Acute intake is more consistent than a course.
The emerging benefit for aerobic endurance and short anaerobic tasks did not survive the sensitivity analyses.
No differences by sex, training status, dose or timing — the popular protocol subtleties are not supported by data.
Certainty on GRADE: low and very low — that means «we don't know», not «it doesn't work».
Standing alongside is a separate 2026 study in 20 participants where citrulline did not improve time to exhaustion on the bike.
A plausible mechanism is no substitute for a result: an elegant biochemical chain is a hypothesis, not proof.

Sources: «Effects of Citrulline Malate Supplementation on Exercise Performance: A Systematic Review and Three-Level Meta-Analysis», Nutrients, 2026;18(12):1881. https://doi.org/10.3390/nu18121881. «Citrulline did not improve high-intensity cycling performance», Examine research feed, 2026. https://examine.com/research-feed/study/15y6M1/