Cannabis does not make CBD, THC or CBC. It makes CBGa, and three enzymes then compete to cyclize it into CBDa, THCa or CBCa. Which of those enzymes a plant carries in working form is what fixes its chemotype before a single flower is harvested.
CBDA synthase was purified in 1996 and described as a single 74 kDa polypeptide with an isoelectric point of 6.1. It needs no coenzyme, no molecular oxygen, no hydrogen peroxide and no metal ion, which rules it out as an oxygenase or a peroxidase. It will accept cannabinerolic acid as well as CBGa, but the turnover number is 0.19 per second for CBGa against 0.03 per second for the alternative, which is why CBDa in a real plant comes overwhelmingly from CBGa.
That settles what a CBGa figure on a certificate means. CBGa is unconsumed substrate, not a separate ingredient the plant adds. Varieties with a defective synthase accumulate it because nothing draws it down, and an early harvest leaves more behind. Heat then separates the paths again, turning residual CBGa into CBG and CBDa into CBD.
CBGa’s strongest documented activity is not at a cannabinoid receptor at all. Screened against store-operated calcium entry in several immune cell lines, the acid cannabinoids outperformed CBD and THC, and CBGa was the most potent of the set, blocking calcium release-activated calcium currents and reducing NFAT activation and interleukin 2 production in human T lymphocytes. A companion study found CBGa the strongest cannabinoid inhibitor of the TRPM7 channel, and showed the effect requires TRPM7’s kinase domain to be functional.
CBDa’s standout result sits at 5-HT1A, where it increased the maximum effect of a 5-HT1A agonist on GTP binding in rat brainstem membranes. The animal work that followed used doses far below anything used for CBD: vomiting in shrews was reduced at 0.1 and 0.5 mg per kilogram, and conditioned gaping in rats at 0.01 and 0.1 mg per kilogram. Both acids also bind the SARS-CoV-2 spike protein with micromolar affinity.
Every one of those findings is a cell line or an animal model, and that is the limit of what can be said. Both acids are pharmacologically interesting, and neither is pharmacologically established in people at any dose.
The asymmetry is worth holding onto. CBDa has the better animal pharmacology and the only human absorption figure of the two, and that figure came from a mixture rather than from CBDa alone. CBGa has the more striking in vitro potency.
In practice this is not a choice between two products. It is a choice about heat. An unheated full spectrum extract carries CBDa as its main cannabinoid with CBGa present in smaller amounts, because both survive when the material is never decarboxylated. A decarboxylated extract from the same biomass gives you CBD and CBG instead.
So you are choosing a profile rather than an evidence base. For the most measured activity in living animals, that is CBDa. For the most potent single in vitro measurement, that is CBGa. For human outcome data, neither qualifies.
Still weighing up the options? See CBGA vs CBG, CBDa vs CBD or CBDA vs THCA. Endoca products related to this comparison: Raw CBD + CBDa oil and CBD/CBDa oil 15%.
Both acids have been treated as lawful when they come from hemp meeting the 2018 Farm Bill limit of 0.3 percent delta-9 THC by dry weight.
From 11 December 2026 that definition changes. Hemp is measured on total THC rather than delta-9 alone, finished products face a limit of 0.4 milligrams of combined total THC per container, and cannabinoids synthesized outside the plant are excluded even where the molecule occurs naturally. Neither acid is THC or converts into it, so neither is caught by the threshold. The exclusion matters more: acids made by conversion rather than extracted would fall outside the new definition. FDA has authorized neither as a food or dietary supplement ingredient.
There is no human safety dataset for either compound in isolation. The single human study of a CBDa-containing product reported adverse events and measurable working memory impairment at its highest dose, and that product averaged 7.2 milligrams of THC, which is the most likely explanation.
Two things follow. An acid form is not automatically gentler than its neutral form, and a raw extract is not automatically milder than a decarboxylated one. Because CBDa reaches far higher plasma concentrations than CBD from the same dose, raw products are where interactions with prescription medicines are least well characterized. Anyone taking medication, pregnant or breastfeeding should speak to a physician first.
Our HPLC panel quantifies acid and neutral forms as separate analytes, which is the only way to prove an unheated extract is actually unheated. A raw batch should show CBDa dominant with CBG and CBD low, and a decarboxylated batch the reverse. Where a cannabinoid falls below the limit of quantification we report it that way instead of estimating it. What a certificate cannot tell you is what either acid does at a receptor.
Yes. CBDA synthase converts CBGa directly into CBDa, with a turnover number of 0.19 per second. The same precursor also feeds THCa and CBCa synthase, so the three acids compete for one pool of CBGa, and which enzymes a plant carries decides its chemotype.
Both, in very different amounts. An unheated CBD-type extract is dominated by CBDa, because the plant’s enzymes converted most of the available CBGa before harvest. CBGa appears in smaller amounts as unconsumed substrate. Heating the same material converts CBDa to CBD and residual CBGa to CBG.
CBDa does, and efficiently. In 15 healthy adults taking a product with equal parts CBD and CBDa, the acid reached peak plasma concentrations 19 to 25 times higher, and reached them sooner. No equivalent human measurement has been published for CBGa, only data in dogs and horses.
Not on the evidence available. Neither acid has a controlled human trial, so there is no outcome to set against CBD, which has three regulatory-grade epilepsy trials behind it. The acids absorb better and behave differently at receptors. Better absorption is not the same as better results.
There is no trial comparing them together with either alone. Raw hemp extracts naturally contain both, alongside other acids.
As the plant matures, enzymes convert most of its CBGA into other acids, mainly CBDA in hemp. Only a small residual amount of CBGA is usually left at harvest.
Taura F, Morimoto S, Shoyama Y. Purification and characterization of cannabidiolic-acid synthase from Cannabis sativa L. J Biol Chem 1996;271:17411-17416.
Taura F et al. Cannabidiolic-acid synthase, the chemotype-determining enzyme in the fiber-type Cannabis sativa. FEBS Lett 2007;581:2929-2934.
Anderson LL et al. Cannabigerolic acid, a major biosynthetic precursor molecule in cannabis, exhibits divergent effects on seizures in mouse models of epilepsy. Br J Pharmacol 2021;178:4826-4841.
Faouzi M et al. Acidic cannabinoids suppress proinflammatory cytokine release by blocking store-operated calcium entry. Function (Oxf) 2022;3:zqac033.
Suzuki S et al. Cannabigerolic acid (CBGA) inhibits the TRPM7 ion channel through its kinase domain. Function (Oxf) 2024;5:zqad069.
Bolognini D et al. Cannabidiolic acid prevents vomiting in Suncus murinus and nausea-induced behaviour in rats by enhancing 5-HT1A receptor activation. Br J Pharmacol 2013;168:1456-1470.
Zagzoog A et al. In vitro and in vivo pharmacological activity of minor cannabinoids isolated from Cannabis sativa. Sci Rep 2020;10:20405.
Elder HJ et al. The pharmacokinetics and pharmacodynamics of a hemp-derived full-spectrum oral cannabinoid product with a 1:1 ratio of cannabidiol to cannabidiolic acid. Cannabis Cannabinoid Res 2025;10:e299-e313.
De Petrocellis L et al. Effects of cannabinoids and cannabinoid-enriched Cannabis extracts on TRP channels and endocannabinoid metabolic enzymes. Br J Pharmacol 2011;163:1479-1494.
van Breemen RB et al. Cannabinoids block cellular entry of SARS-CoV-2 and the emerging variants. J Nat Prod 2022;85:176-184.
Congressional Research Service. Changes to the statutory definition of hemp and issues for Congress. In Focus IF13136, 17 August 2026.
Last reviewed 6 October 2026.
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.