THC: Chemistry, Pharmacology and Research Status

Roughly 94 atoms arranged in a specific tricyclic ring structure. That's all delta-9-tetrahydrocannabinol is, chemically speaking, and yet it's the molecule that reshaped how pharmacologists think about lipid signalling, spawned a whole field of endocannabinoid research, and became one of the most debated substances in Australian medicine. THC chemistry is interesting at the molecular level well before you get to any clinical questions, and this article stays with the molecule, the pharmacology and the state of the evidence. The short version: THC rewards a careful read.

THC chemistry: what the molecule actually is

Delta-9-THC is a terpenophenolic compound. It belongs to the cannabinoid class, which are 21-carbon molecules characterised by a dibenzopyran scaffold. The systematic IUPAC name is (−)-trans-Δ⁹-tetrahydrocannabinol; the "delta-9" tells you the double bond sits between carbons 9 and 10 of the cyclohexene ring. That specific geometry matters. Shift it to between C8 and C9 and you get delta-8-THC, a structural isomer with a distinctly different binding profile. The same logic applies to the other minor isomers discussed later: small changes in where a double bond sits, or how a ring closes, can alter how well a molecule fits a receptor pocket.

Stereochemistry is part of this picture too. The "(−)-trans" in the name describes the three-dimensional arrangement of the ring junction, and receptors are sensitive to handedness. A molecule with the same atoms in a different spatial arrangement can bind poorly or not at all. That's one reason chemists read the full systematic name rather than just the abbreviation.

In the living cannabis plant, THC doesn't actually exist in quantity until heat or UV exposure acts on its precursor. The plant accumulates THCA (tetrahydrocannabinolic acid), a carboxylated form that doesn't bind cannabinoid receptors with meaningful affinity. Decarboxylation, the loss of the carboxyl group as CO₂, converts THCA to active THC. This is why raw plant material behaves so differently from heated preparations. It's a basic chemistry point that gets lost in a lot of popular writing on the subject. If you want the upstream story of how the plant builds these acids in the first place, the piece on cannabinoid biosynthesis in the plant covers it.

The receptor binding picture

THC is a partial agonist at both CB1 and CB2 receptors, the two primary G-protein-coupled receptors of the endocannabinoid system. CB1 is concentrated in the central nervous system, hippocampus, cerebral cortex, basal ganglia, cerebellum, which accounts for the psychoactive effects. CB2 is found predominantly in immune tissues and peripheral organs, though research published in British Journal of Pharmacology has described CB2 expression in CNS microglia as well. For what happens after a ligand docks, including the downstream signalling cascades, see the explainer on cannabinoid receptor signalling.

The "partial agonist" designation is worth unpacking. A full agonist drives a receptor to its maximal response; a partial agonist produces a submaximal response even when every receptor is occupied. THC doesn't activate these receptors to the same ceiling as a full agonist would. Its intrinsic efficacy at CB1 sits lower than that of the endogenous ligand anandamide in some assays, though THC's much higher metabolic stability means it remains at the receptor far longer. The net receptor occupancy over time is higher than you'd expect from the efficacy number alone.

A related term that trips people up is affinity. Affinity describes how tightly a molecule binds; efficacy describes what the receptor does once it's bound. They're separate properties, and a compound can have high receptor affinity with modest efficacy. Mixing the two up is one of the most common errors in popular summaries of cannabinoid pharmacology.

THC also interacts with other targets: TRPV1 ion channels, GPR18, GPR55, and 5-HT₃A receptors have all been characterised in preclinical work. Whether these secondary interactions carry physiological weight at typical exposure concentrations is genuinely unclear; the preclinical binding data doesn't automatically translate to meaningful engagement in a human physiological context. That's a limitation that keeps coming up when you read through the pharmacology literature carefully.

Pharmacokinetics: how the body handles it

Route of administration changes everything for THC, more so than for most small molecules. Inhaled THC reaches peak plasma concentration within minutes, with bioavailability typically cited in the 10–35% range depending on inhalation technique. Oral administration is a different story entirely.

Swallowed THC undergoes extensive first-pass metabolism in the liver, where CYP2C9 and CYP3A4 convert it primarily to 11-hydroxy-THC (11-OH-THC). This metabolite is itself pharmacologically active, arguably more so at CB1 than the parent compound, before being further oxidised to 11-nor-9-carboxy-THC (THC-COOH), the principal urinary marker used in drug screening. Oral bioavailability is variable, typically 4–20%, which explains both the delayed onset and the unpredictability of oral preparations relative to inhalation. The wide range itself is informative: when a figure is quoted as a span rather than a single value, it usually reflects genuine person-to-person and product-to-product variability, not sloppy measurement.

THC is highly lipophilic (octanol-water partition coefficient around 6,000) and distributes extensively into adipose tissue. Terminal half-life is long and concentration-dependent; anywhere from 20 hours in casual users to over 10 days in heavy users due to slow release from fat stores. This is why standard urine immunoassay tests detect THC-COOH for weeks after last use; they're measuring a metabolite that has been slowly re-releasing from adipose tissue, not active THC.

That distinction between detection and activity is one many readers miss. A positive screen says a metabolite is present; it says very little about timing, amount or any effect. THC's adipose distribution is legitimately one of the more interesting pharmacokinetic stories in natural-product chemistry, because it shows how a lipid-soluble molecule behaves differently depending on tissue architecture.

What the research shows, and what it doesn't

The clinical research base for THC has grown substantially since the early 2000s, but it remains thinner than many assume, largely because decades of Schedule I/Schedule 9 restrictions globally created major barriers to controlled human trials. What exists is uneven in quality.

Preclinical models have characterised CB1-mediated modulation of pain signalling, nausea pathways, and intraocular pressure. These are mechanistic observations; they describe what THC does at a receptor level under experimental conditions. Extrapolating directly to human clinical outcomes requires properly powered randomised controlled trials, and the published RCT literature is smaller and more methodologically varied than the preclinical literature. Many trials have small sample sizes, short durations, and heterogeneous patient populations. I'd be cautious about anyone who presents the research as settled in either direction. The gap between the two kinds of evidence is laid out in more detail in preclinical vs clinical research.

There's also the question of THC versus combined cannabinoid preparations. Much clinical research uses plant-derived extracts or standardised combinations of THC and CBD, making it difficult to isolate THC's specific contribution to observed outcomes. The entourage effect, the hypothesis that cannabinoids and terpenes interact pharmacologically in the whole plant, is plausible and has preclinical support, but cleanly attributing effects to THC alone in a full-spectrum preparation is methodologically hard. If you're comparing the two main compounds, the CBD explainer is the natural companion read.

Researchers have also compared THC's binding affinity to structural relatives like THCV and the more recently characterised THCP, a homologue with a longer alkyl side chain that preliminary data suggest binds CB1 with substantially higher affinity. Whether that translates to proportionally stronger physiological effects in humans is not yet established. My own view is that headline affinity numbers get overplayed; a tighter bind in a dish is a hypothesis, not a result.

Australian regulatory status

Under the Poisons Standard (the document that gives legal force to Australia's drug scheduling), THC is a Schedule 8 controlled drug. This is not a technicality; it means possession, supply, and use without authorisation is a criminal offence in every Australian jurisdiction. Schedule 8 carries the tightest regulatory controls of any therapeutic substance in regular clinical use.

Pharmaceutical-grade THC-containing products approved by the TGA and listed on the ARTG may be accessed through two lawful pathways: via an Authorised Prescriber (a medical practitioner with TGA approval to prescribe to eligible patients) or through the Special Access Scheme (SAS-B or SAS-C, depending on product approval status). Both require a valid prescription from a registered medical practitioner. There is no lawful pathway for self-prescribing or obtaining THC products without this framework.

Unapproved or illicitly sourced products sit entirely outside this framework and carry no quality or dosing guarantees. The TGA has published enforcement guidance noting that supply of unapproved cannabinoid products is illegal under the Therapeutic Goods Act 1989, regardless of how those products are labelled or marketed. A label is a marketing artefact, not evidence of what's in the product, and that is worth keeping in mind when reading any claim about a compound's status.

The TGA's own pages are the right place to confirm current details, since scheduling documents get amended. The Therapeutic Goods Administration publishes the Poisons Standard and its guidance directly.

Open questions in the pharmacology

A few things genuinely keep showing up in the literature as unresolved. One is the relationship between CB1 receptor density (which varies substantially between individuals and is downregulated by repeated THC exposure) and variable pharmacodynamic response. Why two people with similar plasma THC concentrations can have dramatically different subjective and physiological responses is not fully explained by pharmacokinetics alone; receptor expression, genetic variation in CYP2C9, and endocannabinoid tone all likely contribute. The broader context is covered in the endocannabinoid system in whole-body physiology.

Another is the developmental question. Preclinical data consistently show that CB1 receptor density and distribution during adolescence differs from the adult pattern, and that THC exposure in adolescent animal models produces different outcomes than adult exposure. Translating that to human risk is complicated by confounding and by the ethical impossibility of conducting controlled exposure studies in minors, but it's a legitimate ongoing research area, not a settled claim.

Structural variants like delta-10-THC and exo-THC are also getting attention as the analytical chemistry of cannabis extracts gets more sophisticated. These minor isomers appear in trace quantities in some preparations and their independent pharmacology is still being characterised. Until that work matures, it's sensible to treat any strong statement about them with scepticism.

Sources

• Medicinal cannabis guidance, Therapeutic Goods Administration (TGA)
• Cannabinoid Receptors and the Endocannabinoid System, StatPearls, NCBI Bookshelf
• The pharmacology and toxicology of delta-9-tetrahydrocannabinol; British Journal of Pharmacology via PMC
• Poisons Standard (SUSMP), Therapeutic Goods Administration (TGA)

Elise Tran, Science writer, cannabinoids & the ECS