Does L-Theanine Work Without Caffeine? What the Dose-Response Data Actually Shows

The two kinds of evidence that keep getting mixed up

Almost every article on L-theanine blurs two very different findings into one claim. Separating them is the whole point of this page.

The first is a neurophysiological marker: alpha-band activity, the 8 to 14 Hz rhythm recorded from the scalp by EEG. It is reproducible at low doses and it is genuine evidence that the compound reaches the brain. The second is behavioural performance: whether a person actually gets more targets right, or responds faster, on a task. That is the thing you care about when you swallow a capsule at 9am. These do not automatically move together. A change in alpha power tells you the brain state shifted. It does not tell you the person performed better.

L-theanine dose-response, from 50 mg to 400 mg

Here is what was measured, at what dose, in the studies that are actually citable. Each row is labelled with which of the two evidence classes it belongs to.

Read down the "result" column and the shape is obvious. The EEG signal is present from 50 mg. The behavioural signal only becomes consistent at 200 mg — and the most-cited 100 mg experiment found nothing at all from L-theanine alone.

What happens when you remove the caffeine

Kelly 2008 is the cleanest test of this question, because it ran all four conditions in the same participants: placebo, 100 mg L-theanine, 50 mg caffeine, and both together. The combination improved both hit rate and discriminability. Caffeine alone improved discriminability but not hit rate. L-theanine alone improved neither.

That is one study at one dose and should not be over-read. But it is the study most often quoted as proof that L-theanine sharpens attention — and what it actually shows is that the combination does. A caffeine-free product, by definition, does not deliver the condition that produced that result. Almost nobody selling L-theanine says so.

The counterweight is the pooled data: across 31 randomised trials, a single 200 mg dose taken 30 to 60 minutes before testing improved choice reaction time with a moderate effect size. That is a solo finding, meta-analysed rather than single-study, and it is the strongest argument that the amino acid does something useful alone.

Where the evidence does not show what people claim

Three limits deserve stating without softening.

Not every cognitive test agrees. A 2025 systematic review and meta-analysis in the Journal of Clinical Medicine pooled five randomised placebo-controlled trials in 148 healthy adults. It found a dose-dependent benefit on rapid visual information processing and recognition visual reaction time (mean difference −15.20 ms, 95% CI −28.99 to −1.41), but no significant effect on simple reaction time (−0.46 ms, 95% CI −15.65 to 14.73) or on the Stroop test (−37.38 ms, 95% CI −86.39 to 11.62). The authors' own conclusion was that the benefits "could not be confirmed by all test methods." Complex, choice-based tasks respond. Simple ones largely do not.

Alpha waves are a marker, not an outcome. Gomez-Ramirez and colleagues found that 250 mg reduced tonic alpha but did not change the anticipatory alpha modulation that tracks where attention is being directed. A brain-state shift is real and interesting. It is not the same as thinking more clearly.

Product doses are often below the tested amount. A 100 mg serving is genuine and has published EEG data behind it, but the behavioural evidence clusters at 200 mg. A formula supplying 100 mg — including the Memopezil formula we cover on this site — sits at the thinner end of that range. Make that judgement with the number in front of you rather than a claim.

Practical notes on taking it on its own

Timing is same-day, not cumulative. Unlike bacopa monnieri, which builds over weeks, L-theanine acts within a single dosing window. Take it before the block of work you want it for, not first thing out of habit. We cover how this interacts with the other ingredients in our timing guide.

It does not build tolerance the way caffeine does. That is the structural argument for a stimulant-free approach, and it is a different claim from "it works better." If your problem is that coffee has stopped delivering, start with why coffee stopped working.

It is the fastest-acting ingredient in most nootropic blends. If a multi-ingredient capsule produces any noticeable same-day effect, L-theanine is usually the reason. The memory-related ingredients work on a far slower timescale, as we set out in our evidence review of bacopa and rhodiola.

The verdict

L-theanine on its own is not a placebo, and the 2026 meta-analysis settles that: a single 200 mg dose measurably improves choice reaction time. But the strongest, most-replicated result in the entire literature remains the caffeine pairing, and a caffeine-free product by definition does not deliver it. What a solo dose reliably delivers is a shift in brain state toward relaxed alertness and, at 200 mg, a modest improvement on complex attention tasks. That is a real but bounded claim, and it is the only one the data supports.

Frequently asked questions