CBG (cannabigerol): definition of the mother cannabinoid

CBG stands for cannabigerol, a hemp cannabinoid whose "ger" syllable comes from the geranyl group in its structure.

According to a 2021 review in the Journal of Pharmacology and Experimental Therapeutics, hemp produces more than 100 different cannabinoids. The review describes CBG as the precursor of the most abundant ones. The label "minor" refers to how little of it remains in the mature plant and says nothing about its importance.

CBG does not get you high, any more than CBD does. After 20 mg in the 2024 trial described below, participants reported neither intoxication nor measurable impairment.

CBG identity card

CBG has the formula C₂₁H₃₂O₂ (316.5 g/mol). Yehiel Gaoni and Raphael Mechoulam described its structure in 1964, the year they also published the isolation of THC.

The identifiers come from the public PubChem database. The 1964 paper has no DOI and is cited by journal and page.

CBG is the precursor of CBD. The plant converts CBGA, the acid of CBG, into CBDA, and heat then turns CBDA into CBD. In the Phytogrammes catalogue reference values surveyed in September 2026, CBD ranges from 6.7 to 14.4% across 72 flowers and CBG from 0.4 to 1.0% across eight. Neither produces the psychoactive effect of THC, and French law applies the same framework to both.

The CBD figures for the catalogue are taken from our survey of the 72 flowers.

"Stronger" than CBD: what has been measured

No trial has compared the two in humans, and the two molecules do not act on their targets in the same way. In cells, CBG is a weak partial agonist at the CB1 and CB2 receptors (Walsh et al., 2021); the Journal of Pharmacology and Experimental Therapeutics review cited above places its affinity for these receptors between those of Δ⁹-THC and CBD. On mouse brain membranes, a 2010 study found CBG activating the α2 receptor at very low concentrations and, at higher concentrations, blocking the 5-HT1A receptor. Potency observed in tissue does not predict the effect of a dose in a person: for a product, all you can check is its CBG content on the batch analysis, a composition figure that says nothing about the effect.

Formats, the tolerability seen in trials and precautions are covered in our article on the measured effects and products of CBG.

Two precursors start the chain: olivetolic acid and geranyl pyrophosphate (GPP). An enzyme called olivetolate geranyltransferase (GOT) joins them into cannabigerolic acid, or CBGA. Three enzymes then compete for that substrate, THCA synthase, CBDA synthase and CBCA synthase, and convert it into THCA, CBDA and CBCA respectively. Which one prevails depends on the variety's genetics, and CBD hemp mostly carries CBDA synthase.

These reactions take place in the glandular trichomes, above all those of the female flowers. GOT has been known since 1998, when Fellermeier and Zenk, in Munich, detected its activity in young hemp leaves. In 2019, a Berkeley team rebuilt the pathway in yeast, from CBGA to THCA and CBDA, using the gene of a previously unknown enzyme with that activity.

Each of these compounds is first made in acidic form. Decarboxylation, driven by heat, removes a carboxyl group and gives the neutral forms CBG, CBD, THC and CBC. In 2016, researchers tracked the reaction between 80 and 145 °C, for up to 60 minutes. CBGA and CBDA lost their carboxyl group at nearly the same rate, about half the rate of THCA. The methods are covered in our guide to decarboxylation, and the acidic form of CBD in our CBDA explainer.

CBN is not part of this chain. The plant does not make it: according to Walsh and colleagues, it forms when Δ⁹-THC breaks down under UV light or heat over time. It therefore does not derive from CBG, despite what some diagrams published online show. Our CBD and CBN comparison explains this ageing process.

While the plant flowers, the synthases convert CBGA nearly as fast as it forms, so only a residue is left at harvest. According to a 2023 clinical toxicology study, CBGA is commonly present at around 1% of dry weight in flowers on sale, and bred varieties can contain about ten times as much. The Phytogrammes catalogue pages checked on 22 September 2026 give, for eight CBD-dominant flowers, a reference profile of 0.4 to 1.0% CBG, with no laboratory certificate published. These declared values refer to CBG, while the 2023 study was about CBGA.

High CBG levels come from genetics. In 2005, a team crossed CBG-rich plants with THC and CBD lines and published the results in Euphytica. They linked the build-up of CBG to two copies of a barely functional version of the converting enzyme, labelled B₀. With that enzyme, CBGA is hardly converted and accumulates. The authors see it as a mutated form of CBDA synthase with greatly weakened affinity or catalytic capacity: these plants come from European fibre hemp populations, and their enzyme still converts a little CBG into CBD.

Resins sold for their CBG exist thanks to this breeding work. You will find the CBG 90µ hash and the other CBG products in the resins, hash and pollen category.

Laboratory work shows CBG acting on several targets. Besides CB1 and CB2, the review by Walsh and colleagues lists α2 adrenergic receptors, at which it is an agonist, 5-HT1A serotonin receptors, at which it is an antagonist, several TRP channels and the nuclear receptor PPARγ. All of these results come from cells and animal models.

Next comes a survey published in 2022: a team questioned 127 American adults online who said they used CBG-dominant products. Without a control group, it describes self-reported use and cannot attribute any effect to CBG.

That leaves three placebo-controlled trials, all small, that tested CBG without combining it with another cannabinoid. A 2023 trial in 40 exercise-trained adults, which mixed CBG, CBD and other ingredients, cannot isolate CBG and is not one of them. Published in Scientific Reports in 2024, the first gave 20 mg of CBG or a placebo, double-blind, to 34 healthy adults who already used cannabis, over two sessions a week apart, monitored by video call. It looked at self-reported anxiety, stress and mood, verbal memory and any signs of intoxication; no intoxication or impairment was found. It was funded by CReDO Science, a company working on CBG.

The second, published in December 2025 in Medical Cannabis and Cannabinoids, followed 63 US military veterans with self-reported sleep problems remotely, 33 on CBG and 30 on placebo, triple-blind: 25 mg a day for two weeks, then 50 mg for two more weeks. On self-reported sleep quality, its primary endpoint, as on quality of life and post-traumatic stress symptoms, it found no significant difference between the groups, and only 35 participants followed the protocol without a major deviation. It was funded by Metta Medical, which markets the tablet tested and employs the lead author.

The third, published online in July 2026 in the Journal of Pharmacology and Experimental Therapeutics, is a laboratory study with rising single doses: 12 healthy adults received 25, 50, 100 and 200 mg of CBG and a placebo placed at random in the sequence, blinded. It found no product-related adverse event and no clear pharmacodynamic effect; subjective effects stayed minimal, apart from a few isolated significant changes, each at a single dose. The authors call for studies of repeated dosing. One of them is affiliated with Charlotte's Web, a CBD company, and the abstract does not state the funding.

The authors of the 2024 trial point out its limits themselves: one modest dose, a non-clinical sample of cannabis users, no physical examination or vital signs recorded and no correction for multiple comparisons. With so few participants and funding from the sector, no promise can be made, and CBG is not a medicine. Epidyolex, the only authorised medicine based on cannabidiol, contains CBD and no CBG. The figures from the 2024 and 2025 trials are given in the article on the measured effects of CBG.

CBG is not listed among French narcotics. A CBG-rich flower or resin is still cannabis, though: it can only be sold if it comes from an authorised variety and contains no more than 0.30% Δ⁹-THC. No CBG product meant to be swallowed is authorised. Status checked on 20 October 2026 against the texts cited.

The consolidated narcotics list published by ANSM, the French medicines agency (26 June 2026), which compiles the order of 22 February 1990 and its amendments, lists "cannabis and cannabis resin" in Annex I and "tetrahydrocannabinols" in Annex III, without naming cannabigerol. Nor does its structure contain the benzo[c]chromene core that defines the family of derivatives classified in 2024.

The plant that contains it is still covered: article R. 5132-86 of the French public health code prohibits cannabis, its resin and products containing them; since February 2022, article R. 5132-86-1 leaves the authorisation of varieties without narcotic properties to a ministerial order, and hemp is only permitted within the framework of the French order of 30 December 2021. That order authorises Cannabis sativa L. varieties listed in the EU or French catalogues whose Δ⁹-THC content does not exceed 0.30%, a cap that also applies to extracts and to products containing them. Through article 3 of its decision of 29 December 2022, the Conseil d'État struck down section II of article 1, which banned the sale of raw flowers and leaves. Our article on the legal THC level explains the threshold.

The same figure appears in EU law: for Common Agricultural Policy payments, Regulation (EU) 2021/2115 only accepts a hemp area if the variety's THC content does not exceed 0.3% (article 4). On 19 November 2020, in case C-663/18, the Court of Justice of the EU ruled that a member state may ban the marketing of CBD lawfully produced in another member state, including CBD extracted from the whole plant, only if the measure is suitable for protecting public health and does not go beyond what is necessary. The ruling concerns CBD only.

Semi-synthetic molecules are a separate matter. ANSM classified HHC, HHCO and HHCP as narcotics from 13 June 2023, and added H4CBD, THCP and other derivatives from 3 June 2024. A product sold as "CBG" that contained one of them would come under those decisions.

Swallowed, CBG falls under the novel food rules. The EU Novel Food Catalogue names it in its "Cannabinoids" entry (updated 14 January 2026): extracts of Cannabis sativa L. and products containing non-psychoactive cannabinoids require prior authorisation under Regulation (EU) 2015/2283. Yet the Union list of authorised novel foods, consolidated on 10 August 2026, contains no cannabinoid. On 20 May 2026, the French Ministry of Agriculture announced that its food directorate, the DGAL, would extend checks to all foods that contain or promote CBD "or any other cannabinoid", food supplements included. See our article on why CBD gummies were withdrawn and our overview of CBD regulation in 2026.

A certificate of analysis gives CBG as a percentage by mass for flowers and resins, or in mg/ml for oils, often alongside CBGA. Simply adding the two inflates the total: when it loses its carboxyl group, CBGA (360.5 g/mol) becomes CBG (316.5 g/mol) and keeps only 316.5 / 360.5 = 0.878 of its mass. The formula is therefore total CBG = CBG + 0.878 × CBGA (for CBD, the factor is 0.877). A flower with 0.2% CBG and 0.5% CBGA, for example, holds at most 0.64% CBG once heated.

This total remains a theoretical ceiling: in the 2016 study cited above, the decarboxylation of CBGA and CBDA came with losses the authors could not explain, whereas that of THCA ran without side reactions.

Check three other lines as well: a batch number matching the product sold, Δ⁹-THC at 0.30% or less, and no HHC, H4CBD or THCP if the lab looked for them. Each line is explained in our guide to the certificate of analysis, and our laboratory page lists the catalogue products.