From a 1949 isolation in Kyoto to a contested EEG literature
L-theanine (γ-glutamylethylamide) was first isolated from gyokuro by Yajiro Sakato at Kyoto University in 1949, then reported in Camellia sinensis more broadly through the 1950s. For the next four decades it sat in agronomy journals as a marker of tea quality — the compound whose concentration tracked shade, nitrogen and the first spring flush. Its move into neuroscience came later, after rodent work in the 1990s confirmed it crossed the blood-brain barrier and modulated glutamate, GABA and dopamine signalling at modest doses.
That second life is where most of today’s claims live, and where most of the confusion begins. The substance behaves nothing like a stimulant. It does not wake you up. What controlled studies have measured, in healthy adults, is a small shift in attention tasks and a small change in alpha-band EEG activity when theanine is consumed with caffeine — typically at doses of 97 to 200 mg theanine alongside 40 to 100 mg caffeine. The combination, not theanine alone, is the consistent signal. Our companion piece The theanine-to-caffeine ratio — what the studies measured walks through the specific trials (Owen 2008, Giesbrecht 2010, Kelly 2008) and the doses they actually used — which are often higher than what a single cup of tea delivers.
This is the gap worth naming. A standard 2.5 g serving of Tài Píng Hóu Kuí (太平猴魁) from Hou Keng village in Anhui, brewed Western-style, yields roughly 15 — 25 mg of theanine in the cup. Shaded teas like Japanese gyokuro or the rare Chinese shaded varietals run higher — perhaps 30 — 45 mg per serving — because shading suppresses the conversion of theanine to catechins in the leaf. Either way, you are well below the bolus doses used in EEG studies. Whether the lower, repeated exposure across a slow morning of gōng fū brewing produces a comparable effect is genuinely unknown. It has not been measured.
What we can say is that the traditional Chinese category of qīng xīn (清心) — “clearing the heart-mind,” the calm-alert state associated with morning green tea — describes something the population has reported for roughly a thousand years. Lu Yu’s Chá Jīng (760 CE) is not a pharmacology text, but it documents the felt experience that modern theanine research is attempting, with mixed success, to operationalise. For the producer side of that story — how shading, plucking standard and cultivar shift theanine content — see the regional notes at puerh.app and the cultivar profiles at thetea.app.
The honest position for a wellness encyclopedia in 2024 is this. Theanine is real, the molecule is well-characterised, and the combined theanine-caffeine effect on sustained attention has been replicated enough times to take seriously. The leap from that to “drink this cup and focus better” is not supported. Dose matters. Timing matters. Individual caffeine sensitivity matters more than either. Read the trials, not the supplement labels — and treat tea as tea, not as a delivery vehicle for a single compound.