Reading the L-Theanine Evidence: Part Five: Does the Science Match the Hype?

Abstract

In February 2025, Roderick Dashwood and Francesco Visioli published a review in Nutrition Research whose title asks the question this series has been circling: does the science match the hype for brain health and relaxation?

Their answer is no. Not that L-theanine is inert, or unsafe, or fraudulent — but that the evidence is far too preliminary to support the claims made for it, and that they would urge caution in its use at pharmacologic doses by the wider population.

This part reads their argument, checks it against the twelve other papers in this series, and then does the thing we have been deferring for four parts: states what we think, about a compound we sell.

1.The Distinction the Whole Argument Rests On

Dashwood and Visioli make one separation early and hold it throughout, and it is the most useful idea in the literature.

It is critically important, they write, to distinguish L-theanine as a natural component of tea from its use as a nootropic nutraceutical at pharmacologic doses.

The numbers make the distinction concrete. Brewed tea delivers between 5.8 and 32 mg of L-theanine per 200 ml cup. One survey of commercial teas found that, contrary to popular wisdom, a standard cup of black tea contained the most at 24.2 mg plus or minus 5.7, while green tea contained the least at 7.9 plus or minus 3.8. Across 37 commercial varieties, average content was 6.56 mg per gram; another survey found concentrations ranging from 0.07 mg per gram in pu-erh to 33.37 mg per gram in white tea.

The authors' conclusion from this spread is blunt: the L-theanine content of brewed tea appears to be non-standardisable and varies markedly according to cultivar and brewing method. They note, dryly, that this variability is itself part of what drove the development of supplements — where quality control can at least be specified.

So the millennia of safe tea consumption, which is the first thing any marketing copy reaches for, describes exposure to perhaps 20 mg at a time. The trials in Parts One to Three used 50 to 450 mg. The gut-brain mechanism in Part Four required a human-equivalent of around 6 grams. These are three different things wearing one name.

2.What the Review Concedes

It is not a debunking, and the parts it accepts are worth stating clearly.

The pharmacology is real. L-theanine is well absorbed, crosses the blood–brain barrier via the leucine transport system, and reaches the brain within an hour with levels elevated for around five. In humans, 100 mg produces a plasma peak of about 25 micromolar at 0.8 hours. It binds glutamate receptors, acts as a partial agonist at the glycine site on NMDA receptors, and has been identified as a competitive inhibitor of the cannabinoid CB1 receptor.

The receptor affinities also come with a number that rarely accompanies the mechanism in popular accounts: at kainate and AMPA receptors, L-theanine is reported as 80- to 30,000-fold less effective than glutamate itself.

The safety is real and unusually well characterised. No toxicity in rats at 6,500 mg per kilogram for two weeks or 2,000 mg per kilogram for 28 days; a no-observed-adverse-effect level of 4,000 mg per kilogram, the highest dose tested; no carcinogenicity at 5 per cent of the diet for 78 weeks; no genotoxicity in the Salmonella assay. On safety grounds this is among the better-documented compounds in its category.

And the alpha-wave literature is real, if narrower than usually reported. Standard supplement doses of 50 to 250 mg increased alpha waves in healthy people — but the associated relaxation was observed only in people with slightly higher than average basal anxiety, or during periods of stress. The review's synthesis of the divergent findings is the phrase calm attention: alpha activity is associated both with relaxation and with selective attention, which is why the same signal supports two apparently opposed claims.

3.Where It Says the Evidence Fails

On anxiety, which is the application most confidently marketed, the review is unambiguous. A comparative study of 200 mg L-theanine against 1 mg of alprazolam found that neither significantly reduced symptoms in an anticipatory anxiety model. Another found no difference from placebo at that dose. A systematic review and network meta-analysis of herbal treatments reported that L-theanine did not outperform placebo, in contrast to other agents in the same analysis.

The consistent qualifier is baseline state: benefit appeared in people with high baseline anxiety, and those who were not anxious failed to derive benefits beyond placebo. That is a meaningful boundary condition and it is almost never stated on a label.

On sleep, the review's position is stronger than anything in Part Two: there are, to date, no gold-standard randomised placebo-controlled trials examining the effects of L-theanine on sleep induction and maintenance. The competing systematic reviews we compared were assembling trials that this review does not consider adequate to the question.

On memory and attention, the two studies cited most often used a supplement containing 360 mg of green tea extract with 60 mg of L-theanine — a combination, in people with mild cognitive impairment, for 16 weeks. The same structural problem as Parts Three and Four.

And then the observation that, arriving independently, confirms something we noticed while reading the mechanistic papers: many studies on the biological properties of L-theanine come from one geographical area, and from one commercial or proprietary source of L-theanine, with data published in journals of moderate impact.

In Part Four we noted the same industry scientist appearing on both a Japanese cerebellar study and the human sleep trial that anchors the sleep evidence; a maze built by the ingredient supplier; disclosure statements that read no competing interests above affiliations saying otherwise. Two independent readings, one conclusion. A literature that looks like replication may be a literature that is closely held.

4.The Regulatory Answer Already Exists

One sentence in the review deserves separating out, because it settles a question that consumer marketing treats as open.

A health claim for theanine and attention was submitted for evaluation in the European Union and denied. Dashwood and Visioli reference the guidance underlying that decision and state plainly that they concur with the assessment.

This is why L-theanine appears in European products with no claim attached, and why the copy around it works so hard — describing tea ceremonies, monks, calm focus, ancient wisdom — while never quite saying the thing a claim would say. The absence of a claim is not an oversight. It is a decision that was made, on the evidence, and lost.

It is worth contrasting with what the same regulator did accept. As the Aberdeen meta-analysis in Part One notes, the European Food Safety Authority gave a positive evaluation to a claim for black tea and attention — and based that assessment on the caffeine content.

5.The One Combination That Survives

Across five parts and thirteen papers, one finding has held its shape under every kind of scrutiny, and it is not an effect of L-theanine.

It is what happens when theanine is added to caffeine. The review's account: 150 mg of caffeine improved fatigue perception, rapid visual information processing and reaction time; adding 250 mg of L-theanine preserved those benefits while enhancing alertness, improving reaction time, and reducing the incidence of headache — which was increased in the caffeine control group. At lower doses, 50 mg of caffeine with 100 mg of theanine improved reaction time and task switching, with fewer effects from distracting stimuli.

A separate crossover study found that the combination circumvented the vasoconstrictor effects of caffeine, without theanine having vascular effects of its own.

The Aberdeen meta-analysis found the pooled effects for theanine plus caffeine on attention switching and digit vigilance accuracy, not for theanine alone. The 2021 systematic review found the combination outperforming either compound individually on selective attention. Even the ADHD study buried in that review found that caffeine and theanine each worsened inhibitory control on their own, and only the pair improved it.

The honest description of this compound, after reading everything above, is that its best-supported role is modifying what caffeine does — smoothing its edge, and possibly making it more tolerable — rather than doing anything substantial by itself.

That is a much smaller claim than the market makes. It is also, unlike most of what the market makes, supported.

Editorial Comment

This is the point in a five-part series where a company that sells the ingredient explains why the evidence, correctly understood, was about its product all along. We are not going to do that, so here is the position plainly.

L-theanine has no authorised health claim in the European Union. A claim for theanine and attention was assessed and rejected. We make no claim for it, on this site, on our packaging or anywhere else, and we agree with Dashwood and Visioli that the science does not currently justify one.

L-theanine is in Clarity, and in Boost at 250 mg with 100 mg of caffeine. Reading these thirteen papers did not change those numbers, and we want to be exact about why. The 250 mg figure and the pairing come from the caffeine-combination literature in section 5 — the only part of this field that survived five parts of reading. That literature describes acute effects on laboratory attention tasks, measured in milliseconds, at low certainty of evidence. It is a reason to design a formula a particular way. It is not a promise, and we have not made one.

What we would ask a reader to take from the series is not a verdict on one amino acid. It is four questions that worked on every paper here and will work on the next one: What else was in the capsule? What did the pre-specified primary outcome do? Who paid, and who is listed above the disclosure line? And is the dose one a human would take?

Those questions cost us something here — a sleep claim, a stress claim, an immunity claim, a gut-brain claim, all of which we could have written and none of which we can defend. We would rather publish that than the alternative, and a reader who applies the same four questions to our formulation pages is doing exactly what we would want.

One practical note to close on. The compound has an unusually clean safety record — no observed adverse effects at doses far above any supplement, no genotoxicity, no carcinogenicity signal — and if you take it and find it suits you, nothing in this series is a reason to stop. But if you are taking it for anxiety that has become persistent, for sleep that has stopped working, or for stress that no longer lifts, the evidence above suggests you are unlikely to get what you need from it, and those are all things worth taking to a doctor instead.