The acidic precursor the entire cannabinoid family is built from — and the literal reason CBG is called the mother cannabinoid. Established as biochemistry; its own pharmacology is still being catalogued.
In the plant, cannabigerolic acid (CBGa) is the single precursor. Formed from olivetolic acid and geranyl pyrophosphate, it is branched by three synthase enzymes into the THCA, CBDA, and CBCA lines — and, on decarboxylation, becomes neutral CBG. CBG is literally the decarboxylated form of CBGa.
Established biochemistry — the branch enzymes have been cloned and characterized, and decarboxylation kinetics measured. Each is traceable to a verified study.
In plain terms: CBGa is the plant's raw material — enzymes convert it into the THC, CBD, and CBC families, and gentle heat turns it into CBG. Switch to Researcher depth for the enzyme-level detail and citations.
CBGA is the substrate of THCA synthase, which oxidatively cyclizes it to THCA. The enzyme has been molecularly cloned and heterologously expressed.
CBDA synthase — purified and characterized from Cannabis sativa — converts the shared CBGA substrate to CBDA.
Structure–function analysis of the branch synthases acting on CBGA — including evidence that they are not perfectly specific, which shapes the cannabinoid a plant makes.
With heat and time CBGA loses CO₂ to form neutral CBG, following temperature-dependent first-order kinetics — why finished products can contain CBG derived from CBGA.
CBGa's own receptor pharmacology is far less characterized than that of its products, and CBG Atlas still asserts no target profile for CBGa. But “in progress” is not “blank”: individual direct findings for CBGa are published and verified, and they are set out in direct pharmacology below — ungraded, and cited as such. The characterized receptor profile still belongs to its decarboxylated form, CBG.
Distinct from the biosynthesis above: these are findings about CBGa acting on something, rather than CBGa being converted into something. Both are verified against their primary sources; neither is a registry-graded claim, and both are cited here with that stated.
In plain terms: researchers have begun measuring CBGa directly rather than only as a precursor. So far that work is in cells and in mice — one ion channel it blocks, and one mouse seizure study whose result went in both directions, helping in some tests and worsening seizures in others. None of it shows what CBGa does in a person, and none of it is a reason to take anything. Switch to Researcher depth for the findings and their citations.
Across a panel of major and minor cannabinoids, about half suppressed TRPM7 currents to some degree and CBGa was the strongest inhibitor of the set — the most fully characterized direct action reported for CBGa in the sources catalogued here. The inhibition requires a functional TRPM7 kinase domain, is sensitized by intracellular Mg·ATP and by free magnesium, and is reduced when intracellular calcium rises. CBGa also inhibited native TRPM7 currents in a B-lymphocyte cell line.
What this is not: A cell-level channel mechanism, and nothing more. It is not an anti-inflammatory, anticancer or kidney finding, it is not a treatment claim, and it has not been tested in people.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
Screening the cannabis plant for anticonvulsant phytocannabinoids identified CBGa as the most potent of those found — and its effects were divergent within the same study. CBGa was anticonvulsant against hyperthermia-induced and spontaneous seizures and in the maximal-electroshock threshold test, and it potentiated clobazam; it was proconvulsant in the 6-Hz threshold test, and a high dose increased spontaneous seizure frequency in the Scn1a mouse model. Surveyed against multiple epilepsy-relevant targets, CBGa showed reported interactions including GPR55, TRPV1 channels and GABA receptors.
What this is not: Mice, not people — and the result cuts both ways inside one paper. It is not evidence that CBGa treats epilepsy or any seizure disorder. The proconvulsant finding is part of the result, not a footnote to it, and must never be dropped to leave the anticonvulsant half standing alone.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
CBGa is the acid; CBG is what it becomes when it loses a carboxyl group. The honest contrast:
| Property | CBGa | CBG |
|---|---|---|
| Chemical form | Acidic (carboxylic acid) | Neutral (decarboxylated) |
| Molecular formula | C₂₂H₃₂O₄ · 360.49 | C₂₁H₃₂O₂ · 316.48 |
| Biosynthetic role | The precursor — branched into THCA/CBDA/CBCA and, on decarboxylation, CBG | A product — the decarboxylated form of CBGa |
| Conversion | Loses CO₂ (heat/time) → CBG | Formed when CBGa decarboxylates |
| Characterized pharmacology | Early — TRPM7 ion-channel inhibition in vitro (kinase-dependent); broader target profile catalogue in progress | α2-adrenoceptor, 5-HT1A, CB1, TRP channels (in vitro) |
| Human evidence | No trial of isolated CBGa. What exists is exposure and co-presence — CBGa measured in human samples after hemp-product intake, and present in a product dosed to trial participants. Neither is efficacy. | One small acute trial on anxiety and stress (Low); other human CBG research exists |
| Intoxicating | No | No |
“No human evidence” is the easy line, and it is wrong. Three different things are true at once — CBGa has been measured in people, CBGa has been given to people inside a product, and CBGa has never been trialled on its own. Keeping those apart is the whole of the honest answer.
In an ethics-committee-approved single-dose administration study, participants consumed commercially available hemp products and urine collected before and after consumption was analysed with a validated panel of sixteen cannabinoids — CBGa among the analytes. Variable patterns of cannabinoids and their metabolites were found in those samples.
What this does not establish: This is analytical chemistry: it establishes that CBGa is a real analyte in human samples after ordinary hemp-product intake. It measures no effect on any symptom or condition. The published abstract does not itemize per-cannabinoid urinary results, so no CBGa concentration is stated here — and none should be added without the primary source in hand.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
The only human trial of CBG's effects on anxiety and stress administered a tincture that, by its own Methods section, also contained CBGa alongside CBG and β-caryophyllene. Participants in that trial therefore received CBGa — as a minor constituent of a CBG product, not as the thing being tested.
Verbatim, from its Methods: “The CBG tincture was composed of 10 mg/ml CBG, 0.89 mg/ml CBGA, 0.35 mg/ml β-caryophyllene.”
What this does not establish: That trial measured CBG, and its outcomes are NOT attributable to CBGa. CBGa was neither the intervention nor an isolated variable, no arm of the study varied it, and nothing in that trial grades CBGa. It is cited here for one fact only — that CBGa has been present in a product administered to people under trial conditions.
A graded entry in the CBG Atlas registry — but graded for CBG, not for CBGa. It is cited here for its published composition only, so it carries no CBGa evidence grade and supports no CBGa efficacy claim.
No trial of isolated CBGa in people has been published: no efficacy trial, and no Phase-1 safety or dose-ranging study of CBGa on its own. This is the tier that carries the weight of the honest answer, and it is unchanged.
What this does not establish: This is an absence of testing, not a negative result. CBGa has not been trialled in people and found wanting — it has not been trialled. Nothing here is evidence either for or against an effect, and it must never be read as either.
Tier 3 is an absence of testing, not a verdict. CBGa has not been trialled on its own in people and found wanting — it has not been trialled on its own. That is a gap in the literature, and CBG Atlas records it as a gap rather than filling it in either direction. If an isolated-CBGa trial is published, it will enter through the same gated pipeline as everything else on this page, and this tier will change.
CBGa's registry-graded evidence is its biosynthesis — the studies characterizing how it becomes the other cannabinoids. That graded set is below. The direct-pharmacology and human-record sources earlier on this page sit outside it: published and verified, cited in prose, and carrying no grade here.
CBGa is non-intoxicating. No therapeutic effect of CBGa is established in humans — because no trial of isolated CBGa has been run. That is an untested question, not a tested-and-failed one. Human data on CBGa does exist, and it is exposure and co-presence data rather than efficacy data: see the human record. Its significance on this site remains as the precursor: on heating it converts to CBG, so a finished product's CBG can originate from CBGa. Any cardiovascular caution (the α2-adrenoceptor blood-pressure note) applies to CBG, its decarboxylated form. This is educational information, not medical advice.
The verified studies behind this page — each links to its PubMed record. The graded registry set comes first; the sources cited in prose, which carry no grade, are listed separately below rather than blended in.
Each of these was verified against its PubMed record before it was cited. None of them is a graded CBGa record in the registry export this page reads, so none carries an evidence grade here — and every place they appear above says so.
Analytical exposure only. It shows CBGa is an analyte measured in human samples after controlled hemp-product intake. It measures no effect on any condition and supports no efficacy claim.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
Cited for ONE fact from its Methods: the administered tincture also contained CBGa. Its outcomes are a CBG result and are not attributable to CBGa, which was neither the intervention nor an isolated variable.
A graded entry in the CBG Atlas registry — but graded for CBG, not for CBGa. It is cited here for its published composition only, so it carries no CBGa evidence grade and supports no CBGa efficacy claim.
Cell-level channel mechanism only. Not an anti-inflammatory, anticancer or kidney claim — the therapeutic-potential sentence in its own abstract is the authors' speculation and is not reproduced here.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
Preclinical, in mice, and DIVERGENT within the one paper — anticonvulsant in some models, proconvulsant in another. Never citable for a human epilepsy or seizure claim, and never citable for the anticonvulsant half alone.
Verified against its PubMed record and cited in prose. This paper is not yet a graded entry in the CBG Atlas registry, so it carries no evidence grade here.
Seven compounds, each with its own monograph. The line under each one is what the verified registry actually holds for it — including where that is nothing yet.
The mother cannabinoid. CBGa decarboxylates to CBG, and both occur in the plant — it is CBGa that the three synthases branch into THC, CBD and CBC. The best-evidenced compound on this platform.
The propyl (varin) analogue of CBG — a three-carbon side chain instead of five.
Non-intoxicating, and one of the three branch products off the shared CBGA substrate.
Forms as THC ages and oxidises, so it accumulates in stored material rather than being made by the plant directly.
The propyl analogue of CBD, and the varin with the largest published human trial record.
The propyl analogue of THC, with receptor behaviour that differs from it in laboratory work.