ORIGENHuman Potential MeasuredYour protocol with Oura or WHOOP
A quiet pre-sleep ritual scene.

Sourced, not promised.

Research / Mental Work

How strong is the evidence for L-theanine and falling asleep?

By Alejandro del Palacio · · 5 min read

The short answer

Weak for falling asleep, and self-reported. The Bulman 2025 meta-analysis (Sleep Medicine Reviews) found an SMD of 0.15 (confidence interval 0.01 to 0.29; 10 studies) in how long people take to fall asleep, as reported in questionnaires. It is a very small effect, expressed in standard deviations, not minutes. A ring may not see it.

Key findings

  • The Bulman 2025 review gathered 19 articles, 18 of them in the meta-analysis (Sleep Medicine Reviews).

  • On self-reported sleep onset latency, the pooled effect is SMD 0.15 (confidence interval 0.01 to 0.29; 10 studies). The authors note that studies on "pure" L-theanine are lacking.

  • The 18 trials mix healthy people with people who have sleep complaints.

  • StatusEstimated

The myth: “L-theanine makes you fall asleep fast”

Short answer: the evidence reviewed here does not show it. In the Bulman 2025 meta-analysis, the effect on how long people take to fall asleep was SMD 0.15 (confidence interval 0.01 to 0.29; 10 studies): a very small effect, measured with questionnaires.

In this article
  1. What did the 2025 meta-analysis find on L-theanine and sleep?
  2. How big is an effect of 0.15, and how was it measured?
  3. What can you measure with your ring, and what can't you?
  4. Where the evidence ends

What did the 2025 meta-analysis find on L-theanine and sleep?

It found a difference in how long people say they take to fall asleep, not in what a device records.

The work is by Bulman and colleagues, and it was published in 2025 in Sleep Medicine Reviews. It is a systematic review with a meta-analysis.

The review included 19 articles, and 18 went into the meta-analysis. Those 18 are randomized trials and they mix healthy people with people who have sleep complaints. So the result does not describe a single type of reader.

The figure we care about here is sleep onset latency: how long you take to fall asleep once you go to bed. That figure pools 10 studies. It is the only effect figure we use here.

The result was an SMD of 0.15. It is the pooled difference in the latency people reported.

How big is an effect of 0.15, and how was it measured?

It is a very small effect, and latency was measured by questionnaire.

The SMD (standardized mean difference) expresses the difference in standard deviations, not in minutes. By the definition of the measure, a value of 0.15 equals 0.15 standard deviations: a small fraction of the variation between people.

The confidence interval runs from 0.01 to 0.29. It is the range of values compatible with these data. The low end, 0.01, sits almost on zero. If the true effect were that, it would be 0.01 standard deviations.

The outcome is self-reported sleep onset latency, by questionnaire. There is no polysomnography behind the figure: no lab recording confirms that latency.

A questionnaire records what a person perceives. It is not a device measurement.

ItemWhat the source saysStatus
SourceBulman 2025, Sleep Medicine ReviewsEstimated
DesignSystematic review and meta-analysis: 19 articles, 18 in the meta-analysisEstimated
OutcomeSelf-reported sleep onset latency (questionnaire)Estimated
EffectSMD 0.15, confidence interval 0.01 to 0.29; 10 studiesEstimated
An empty concrete terrace at blue hour facing misty mountains and a still lake.

What can you measure with your ring, and what can't you?

If you use Oura or WHOOP, you can measure your own sleep latency. What you cannot expect is for your device to reproduce what the meta-analysis measured.

The meta-analysis pooled self-reported latency, not latency measured by a device. So a ring, on its own, does not settle this question. Even if the effect is real, a device may not see it.

If you have worn your device for a while, you already hold a series of your own latency. That series tells you how much your nights vary when you change nothing. With that reference, a change of yours either stands out from the noise or stays inside it.

This piece does not recommend taking it: it describes what each source measures.

Related reading:

A person seated on a stone bench at dusk with a steaming cup of tea beside them, facing a still lake and misty mountains.

Where the evidence ends

This data point is narrow. Here is what it does NOT say:

  • The 0.15 is subjective latency only, from questionnaires, in 10 studies. It does not extend to device-measured sleep quality.
  • There is no polysomnographic confirmation and no size in minutes.
  • There is no verified dose: the article's abstract does not give one.
  • We did not extract the total number of participants here.
  • It does not say it works for you or that a ring will detect it.

What it means

For falling asleep, the evidence is thin and subjective: an SMD of 0.15 in 10 studies. The authors note that studies on "pure" L-theanine are lacking. If you want to know what happens in your case, the question is answered with your own latency series, not with the average of a meta-analysis.

Quick questions

What did the 2025 meta-analysis find on L-theanine and sleep?

It found a difference in how long people say they take to fall asleep, not in what a device records. The work is by Bulman and colleagues, and it was published in 2025 in Sleep Medicine Reviews. It is a systematic review with a meta-analysis. The review included 19 articles, and 18 went into the meta-analysis.

How big is an effect of 0.15, and how was it measured?

It is a very small effect, and latency was measured by questionnaire. The SMD (standardized mean difference) expresses the difference in standard deviations, not in minutes. By the definition of the measure, a value of 0.15 equals 0.15 standard deviations: a small fraction of the variation between people.

What can you measure with your ring, and what can't you?

If you use Oura or WHOOP, you can measure your own sleep latency. What you cannot expect is for your device to reproduce what the meta-analysis measured. The meta-analysis pooled self-reported latency, not latency measured by a device. So a ring, on its own, does not settle this question.

Sources

  1. Bulman 2025, Sleep Medicine Reviews (systematic review and meta-analysis). Article (DOI) · PubMed

Alejandro del PalacioAuthor · reviewed on 2026-11-09

Measurement tool, not medical advice.

ORIGENHuman Potential Measured

pieza_id p1-l-theanine-sleep-evidence