How Long Does L-Theanine Take to Work? The Pharmacokinetics, Then the Evidence

Short answer: L-theanine appears in the blood about 10 minutes after you swallow it, peaks at roughly 50 minutes, and has an elimination half-life of about 65 minutes, so most of a dose is gone within four to five hours. The trials that measured a real effect gave 200 mg, 30 to 60 minutes before the task.

That is the timing question answered. The more useful question is what happens at minute 50, and there the honest answer is narrower than almost anything written about this supplement.

The pharmacokinetics

Only one human disposition study is usually cited, and it is worth reading closely. Van der Pijl and colleagues gave male volunteers 25 to 100 mg of L-theanine, as ordinary tea, as theanine-enriched tea, and as a plain aqueous solution.

Two consequences follow directly and neither is usually stated.

Taking it "an hour before" is right, and it is not an arbitrary round number. Fifty minutes to peak means a dose taken with the intention of covering a meeting or a task should go in roughly an hour ahead.

It does not last the evening. With a 65-minute half-life, a 200 mg dose is down to about 25 mg after four hours and effectively gone by five. Anything you notice six hours later is not the theanine still circulating. This matters for the sleep claim, below.

The exposure was dose-proportional, and taking it in tea produced essentially the same curve as taking it in water, so brewing it does not change the timing.

One disclosure belongs with those numbers. The paper was authored by Unilever researchers, and Unilever markets products containing L-theanine. That does not make the pharmacokinetics wrong, and physical measurements of plasma concentration are about as robust as this literature gets. It is worth knowing because the funding pattern turns out to be the story here.

What it actually does at minute 50

The best current synthesis is a 2026 meta-analysis of 31 randomised trials covering 1,168 people. Its findings, in the order the evidence supports them:

Reaction time, yes. A single 200 mg dose taken 30 to 60 minutes before cognitive testing significantly improved choice reaction time, with a standardised mean difference of 0.51 (95% CI 0.25 to 0.77). That is a moderate effect with a confidence interval that does not touch zero. It is the most solid thing in the entire theanine literature.

Acute stress, weakly and suspiciously. The reduction was "modest (SMD = 0.31) and largely influenced by studies with a high risk of bias." When an effect is carried by the weakest studies in a pool, that is the pattern you see before an effect disappears in better trials.

Anxiety, no. "Anxiety effects were inconsistent and non-significant," with a single exception in one trial on psychotic anxiety at 400 mg a day for eight weeks.

Fatigue, no. No significant effect.

An earlier meta-analysis of the tea constituents found moderate effect sizes in the first two hours after dosing for alertness and attention-switching accuracy. Note the exact wording: those effects were "in favor of combined caffeine and L-theanine", not of theanine on its own. But it also reported that its moderator analysis "revealed trends toward greater change in effect size for caffeine dose than for L-theanine dose, particularly during the first hour postdose." The caffeine was doing more of the work than the theanine.

Anxiety effects were inconsistent and non-significant except for one study on psychotic anxiety.

Gerolymos, Saddier, Boyer & Fond, Molecular Psychiatry, 2026

The calm claim has been tested and it did not hold

This is the finding that should change how you read a theanine label. The substance is sold almost entirely on relaxation, and relaxation is the outcome that keeps failing.

European regulators looked at this in 2011 and issued four separate conclusions. For cognitive function: "A cause and effect relationship has not been established," because the three usable human studies "did not show an effect." For alleviation of psychological stress, for maintenance of normal sleep, and for reduction of menstrual discomfort, the reason given was blunter: "No references were provided from which conclusions could be drawn for the scientific substantiation of the claim." No permitted health claim for theanine exists in the EU.

Then in 2018 regulators assessed a black tea attention claim submitted by a manufacturer, whose case rested explicitly on "the concerted action of two substances, caffeine and l-theanine." The panel approved a claim, but not the one submitted. Its conclusion was that the observed effect on attention "can be explained by its caffeine content." The authorised wording is "Owing to its caffeine content, black tea improves attention."

An industry dossier built around theanine synergy was assessed, and the regulator credited the caffeine.

The quality of the evidence, stated plainly

A 2025 systematic review by Payne and colleagues examined 50 randomised trials of tea, theanine, and the combination. Its risk-of-bias assessment is the single most informative sentence available on this topic:

"Overall, zero studies had a low risk of bias across all domains."

Not one clean study in the field.

On funding, the same review found industry "was the most common funding source (n = 8), followed by academia (n = 4), and government (n = 2)," with four studies not reporting funding at all. Its own author list includes employees of a tea company and a food manufacturer.

The samples are also small. That review's pooled trials had a median sample size of 27. The most-cited alpha-wave finding, the origin of nearly every "theanine calms your brain" claim, is a two-page conference abstract in a journal supplement with 16 participants in one arm and 19 in the other, reporting no effect size and a typo where the p-value should be.

A 2025 review in a nutrition journal put the position more directly than most papers do: in the absence of well-designed and carefully controlled human trials, the authors "would urge caution" and concluded that "the science does not yet match the hype behind this trending supplement for brain health and relaxation."

Tea will not get you there

People reasonably ask whether they can skip the capsule. The measured answer is no, not at any sane number of cups.

Keenan and colleagues measured 27 commercial teas brewed for two minutes at 80 degrees:

Black tea contained the most and green tea the least, which reverses the usual assumption. Brewing time was the major determinant of how much was extracted. Milk and sugar in small amounts made no difference; a lot of milk lowered detectable theanine.

By simple arithmetic, and this is our arithmetic rather than a finding in the paper, a 200 mg dose is roughly eight cups of black tea or twenty-five cups of green tea. Even the 50 mg used in the early EEG work is about two cups of black tea.

Which raises the obvious question: if you drink tea for the calm-alert feeling, you are getting a small amount of theanine and a meaningful amount of caffeine, and the regulator has already ruled on which of those is doing the work.

Sleep

Two credible reviews disagree, and the disagreement is instructive.

A 2025 meta-analysis of 19 articles covering 897 participants found small significant improvements: subjective sleep onset latency SMD 0.15 (95% CI 0.01 to 0.29), daytime dysfunction 0.33, and overall subjective sleep quality 0.43. Every significant pooled outcome was subjective, and the review noted that "the lack of studies on 'pure' L-theanine warrants further investigation," because many pooled trials used theanine inside multi-ingredient formulas.

Payne and colleagues, reviewing the same territory in the same year, could not pool the sleep data at all. Eight of their trials reported sleep outcomes and the results were too heterogeneous to combine, so they were described narratively rather than meta-analysed. When two teams look at overlapping literature in the same year and one produces small pooled effects while the other concludes the data will not pool, that is itself a statement about the evidence.

Set against the pharmacokinetics, the mechanism is hard to see. A 65-minute half-life means the compound is gone before the middle of the night. Whatever is being reported on morning questionnaires, it is not theanine still acting at 3am.

What is reasonable to conclude

Theanine is safe at ordinary doses. It has US GRAS status as a food ingredient at up to 250 mg per serving, and the 2026 meta-analysis reported no serious adverse events across 31 trials. Being safe is not the same as being effective, and GRAS is a food-ingredient safety determination rather than an efficacy finding. The FDA has never assessed theanine for anxiety, sleep or cognition. The one regulator that did assess it for exactly those uses, in 2011, rejected every claim.

If you want to try it, the evidence-shaped way to do so is 200 mg, roughly an hour before something demanding, expecting slightly faster reactions and nothing dramatic. If you take it and feel calm, that is a real experience, and it may well be worth something to you. It is simply not a result the controlled trials have been able to reproduce against placebo, and the trials that came closest were mostly paid for by companies selling it.

The reaction-time finding is genuine. The relaxation story is, at present, an industry with better marketing than data.

Questions people ask

How long does L-theanine take to kick in? Plasma levels peak around 50 minutes after an oral dose, with the compound detectable in blood from about 10 minutes. The trials showing an effect dosed 30 to 60 minutes before the task.

How long does L-theanine last? The elimination half-life is about 65 minutes. Counting from the 50-minute peak, roughly an eighth of the dose remains about four hours after you took it, and it is effectively cleared by five.

How much L-theanine should I take? The dose with the best supporting evidence is 200 mg as a single acute dose. Higher chronic doses, 400 mg daily, appear in a small number of trials with weaker results.

Does L-theanine work for anxiety? The 2026 meta-analysis of 31 trials found anxiety effects "inconsistent and non-significant." European regulators rejected the psychological-stress claim in 2011 because no usable evidence was submitted at all.

Can I get enough L-theanine from tea? Not at supplement doses. A 200 ml cup of black tea contains about 24 mg and green tea about 8 mg, so a 200 mg dose is around eight cups of black tea.

Does L-theanine help you sleep? Reviews conflict, and the improvements that were found are small and entirely subjective. The 65-minute half-life makes a direct overnight mechanism difficult to explain.

Should I take L-theanine with caffeine? The combination has the strongest acute evidence in this literature, for alertness and attention switching in the first two hours. But moderator analysis suggests the caffeine dose drives most of the effect, and European regulators attributed the black tea attention benefit to caffeine specifically.

This article summarises published trial evidence for general information and is not medical advice. Supplements can interact with medication and with existing conditions. Speak to a pharmacist or doctor before adding one, particularly if you take psychiatric or blood pressure medication.