L-Theanine Is the One Dose the Shelf Gets Right — Until It Joins a Formula

Dark indigo card: theanine is dosed right on its own bottle — 200 mg, the trial dose — and halved to 100 mg inside every formula. 675 products, NIH database.

After five counts of what supplement labels hide, we owe the shelf one story about an ingredient it mostly gets right. This is that story — and the exact point where it stops being true.

We read 675 products containing L-theanine from the NIH’s public label database: dedicated theanine bottles, sleep formulas, calm-and-stress blends, and caffeine-pairing focus stacks. The database holds about 2,100 theanine records; ours is the 700 most relevant, so a sample, not the whole shelf.

Standalone theanine is the best-behaved label we’ve counted

The reference point: the trial we cite for evening use gave healthy adults 200 mg a day and measured stress-related symptoms and sleep quality over four weeks (PMID 31623400).

Among 196 dedicated theanine products:

  • The median dose is exactly 200 mg.
  • 120 of 196 — 61% — print precisely 200 mg, the studied dose to the milligram.
  • 78% are at 200 mg or above; every single one states its dose — zero hidden.

Across all 675 products, only 10.8% hide the theanine dose inside a blend — against the 38.6% hidden-dose rate we found across sleep-formula ingredients in general. And 23% name the branded form (Suntheanine) outright.

When an ingredient is cheap enough to dose honestly and its trial number is easy to print, the market prints it. Theanine is the proof that the problems we’ve been counting all series are choices, not necessities.

Then it joins a formula, and the dose halves

Here is the same ingredient the moment it stops being the product and becomes a line on someone else’s label:

Context Products Dose hidden Median stated At 200 mg+
Dedicated theanine 196 0% 200 mg 78%
Sleep formulas 132 15% 100 mg 30%
Calm / stress blends 117 10% 100 mg 29%
Focus / energy stacks 110 24% 100 mg 41%

A clean two-fold drop, and our sixth replication of the dilution effect: ashwagandha fell four-fold entering formulas, magnesium 2.5-fold, theanine two-fold. The pattern has now held for every ingredient we’ve checked.

675 theanine products · NIH database

Median theanine dose, by what the label promises

The dark line at 200 mg is the evening-use trial dose. Bars are median stated doses.

dedicatedtheanine is the product 200 mg78% at 200+ · 0% hidden
sleepformulas 100 mg30% at 200+ · 15% hidden
calm / stressblends 100 mg29% at 200+ · 10% hidden
focus / energycaffeine stacks 100 mg41% at 200+ · 24% hidden
standalone — matches the trial dose inside a formula — half of it

Three different promises, one number. And it’s only right once: 100 mg beside ~50 mg of caffeine is the studied focus ratio (PMID 18681988). For evening use the trials used 200 mg (PMID 31623400) — sleep formulas carry half.

Source: NIH Dietary Supplement Label Database, 698 labels read 29 August 2026, 675 listing theanine. Context assigned from product name plus caffeine in the ingredient list. Medians of stated per-serving doses.

The strangest part: three promises, one number

Look at the medians again. Sleep formulas: 100. Stress blends: 100. Caffeine-driven focus stacks: 100. The marketing promise changes completely; the dose doesn’t move a milligram. Whatever is setting these numbers, it isn’t the research — the formulas dose by convention, one convention, everywhere.

And here’s the twist that keeps this honest: in one of the three contexts, the convention is accidentally right. The classic caffeine-pairing trial used 100 mg of L-theanine with 50 mg of caffeine and found the combination improved attention-switching accuracy (PMID 18681988). So a focus stack at 100 mg matches its evidence. A sleep formula at 100 mg carries half the dose its own use case was studied at.

Same number on both labels. One of them is a studied dose. The other is half of one. Nothing on the front of either bottle tells you which is which.

Where we stand, and what this count can’t see

Disclosure, as every time: our formula carries 200 mg of L-theanine — the trial dose — so this study flatters our label twice over. Weigh it accordingly.

The limits:

  • Context comes from the product’s name. We classified by words like “sleep”, “calm”, “focus”, plus actual caffeine in the ingredient list. A product named ambiguously lands in “other” (120 products, median also 100 mg — the convention again).
  • Per serving, as printed. Some products suggest multiple servings a day.
  • Reaching a studied dose is not proof of effect, and 100 mg without caffeine isn’t proof of no effect — it’s territory the trials mapped thinly. The honest reading is narrower: at 200 mg you’re inside the evening-use evidence; at 100 mg you’re inside the caffeine-pairing evidence only if there’s caffeine next to it.
  • The sample is relevance-ranked — 700 of ~2,100 records, 698 retrieved. Deep-catalog products are under-represented.

How to read a theanine label in ten seconds

  1. Buying it for the evening? The trial dose is 200 mg. Standalone bottles almost always have it; formulas usually carry half.
  2. Buying a caffeine stack? 100 mg next to ~50 mg of caffeine is the studied ratio — the one place the shelf’s favorite number is the right one.
  3. No number at all? That’s rare for theanine — one in ten. If the label you’re holding is the exception, that tells you something too.

How we counted, so you can check us

We searched the DSLD for “theanine”, took the 700 most relevant on-market records, and retrieved 698 full labels on 29 August 2026; 675 list theanine in the supplement facts panel and form the sample. Doses were read from the theanine line itself (blends walked, units converted to milligrams); a theanine line inside a blend total with no quantity of its own counts as hidden. Both trial doses are linked to their PubMed records above — verified against the abstracts, not quoted from memory. Duplicate package sizes of the same product appear as separate labels, as the database lists them.

The database is free and needs no account. If you repeat this and get different numbers, we’d like to know.

†These statements have not been evaluated by the Food and Drug Administration. This product is not intended to diagnose, treat, cure, or prevent any disease. This article is educational and is not medical advice; if you have a medical condition or take medication, talk to your healthcare provider.

The opposite end of the spectrum is apigenin: a niche that standardized on 50 mg — a dose with zero human trials behind it.

The Product

Meet Night Reset.

Four actives, printed to the milligram. No melatonin. No proprietary blends. The Whole Protocol Guarantee — 90 nights.

Shop Night Reset — $69

The full protocol — 3 bottles, 90 nights, $1.66 a night