The entourage effect is one of the most cited concepts in cannabis science and one of the most misrepresented. To make evidence-based product decisions, it helps to know what Mechoulam actually proposed in 1998, how Russo extended it in 2011, and what later laboratory and human studies have confirmed, contested or contradicted.
Where the Term Came From
The term “entourage effect” was coined by Shimon Ben-Shabat, Raphael Mechoulam and colleagues in a 1998 paper in the European Journal of Pharmacology. The paper was about endocannabinoids, the body’s own cannabis-like signalling molecules. It found that the activity of 2-arachidonoylglycerol (2-AG) at cannabinoid receptors was enhanced by fatty acid glycerol esters found alongside it. Those esters were inactive on their own. This “entourage” of inactive molecules amplified the active one.
Critically, the 1998 paper was about endocannabinoids, not cannabis plant compounds.
McPartland and Russo (2001, Journal of Cannabis Therapeutics, “Cannabis and Cannabis Extracts: Greater Than the Sum of Their Parts?”) reviewed possible synergy among cannabinoids, terpenes and flavonoids. One of their proposals was that CBD modulates the psychoactivity of THC. In 2011 Ethan Russo published “Taming THC: Potential Cannabis Synergy and Phytocannabinoid–Terpenoid Entourage Effects” in the British Journal of Pharmacology. That paper extended the concept to THC, CBD and terpenes acting together.
Both are narrative reviews. They bring together preclinical findings and propose mechanisms, but they present no original data showing that terpenes change cannabinoid effects in humans. In commercial settings, the claim that “the entourage effect is established science” usually traces back to Russo 2011, cited without saying that it is a hypothesis paper.
Laboratory Evidence: Receptors and Animals
Do terpenes act at cannabinoid receptors? This is contested. Two in-vitro studies were negative:
- Santiago et al. (2019, Cannabis and Cannabinoid Research, “Absence of Entourage”) tested six common terpenes in cells expressing human CB1 or CB2 receptors: α-pinene, β-pinene, β-caryophyllene, linalool, limonene and β-myrcene. The terpenes had no effect on their own and did not change the response to THC.
- Finlay et al. (2020, Frontiers in Pharmacology 11:359, “Terpenoids From Cannabis Do Not Mediate an Entourage Effect by Acting at Cannabinoid Receptors”) used radioligand binding and functional assays. Five terpenes were tested. Apart from a possible weak interaction of β-caryophyllene with CB2, none bound CB1 or CB2, and none changed the activity of THC, CBD or 2-AG.
One study was positive. LaVigne et al. (2021, Scientific Reports, “Cannabis sativa terpenes are cannabimimetic and selectively enhance cannabinoid activity”) tested α-humulene, geraniol, linalool and β-pinene. These terpenes produced cannabinoid-like behaviours in mice and activated CB1 in cell assays. Combined with a synthetic cannabinoid agonist, their effects were selectively additive. These are mouse and cell data, and the terpenes, assays and doses differ from those in the negative studies.
So the receptor question is still open. Terpenes might also act through other targets (TRPV1, serotonin or GABA-A receptors have been proposed), and none of these studies answers that for humans. One terpene–receptor interaction is well documented: β-caryophyllene binds CB2 receptors (Gertsch et al., PNAS 2008).
Whole-plant extract vs. isolate in animals. Gallily et al. (2015, Pharmacology & Pharmacy) compared a CBD-rich extract with purified CBD for inflammation and pain in mice. Purified CBD showed a bell-shaped dose–response, meaning it worked only within a narrow dose range. The extract’s effect kept rising as the dose increased. This has limits: it is a single rodent study, rodent pharmacokinetics differ from human pharmacokinetics, and the non-CBD components responsible were not identified. In the wider preclinical literature, some studies find a whole-plant advantage and others do not.
Human Evidence
Nabiximols (Sativex). Nabiximols is a standardised cannabis extract given as a spray into the mouth. Each 100 µL spray delivers about 2.7 mg THC and 2.5 mg CBD, plus minor plant compounds. Its broad approval, in many countries, is for spasticity in multiple sclerosis (MS). Randomised trials such as Wade et al. (2004) and Collin et al. (2007, 2010) support that use. Pain approvals have been limited. Canada, for example, gave conditional approval (NOC/c) for neuropathic pain in MS (2005) and for advanced cancer pain (2007).
For neuropathic pain in general, the Cochrane review by Mücke et al. (2018) covered cannabis-based medicines, including nabiximols. It found at best a small benefit over placebo. The quality of the evidence ranged from very low to moderate, and adverse effects may limit how useful these medicines are. An earlier meta-analysis by Aviram & Samuelly-Leichtag (Pain Physician 2017; this is not a Cochrane review) concluded on limited evidence that cannabis-based medicines might help chronic pain, mainly neuropathic pain.
Trials that compare nabiximols only with placebo cannot show why the combination works. Any benefit could be:
- additive: two active compounds each doing their own thing;
- pharmacokinetic: one compound changing how the body handles the other;
- synergistic in Russo’s sense.
On the pharmacokinetic option: THC is broken down mainly by the liver enzyme CYP2C9, with some contribution from CYP3A4, and CBD can inhibit these enzymes in vitro. At nabiximols doses, however, the evidence that CBD meaningfully raises THC blood levels is weak. Reviews judge this interaction probably of little clinical importance at low to moderate doses.
The closest direct test: Johnson et al. (2010). One trial did include a THC-only arm (Journal of Pain and Symptom Management). It enrolled 177 people whose cancer pain was not controlled by strong opioids. They were randomly assigned for two weeks to one of three arms: THC:CBD extract (nabiximols), a THC extract, or placebo. Only the THC:CBD extract improved pain significantly more than placebo; the THC extract did not. This is the most direct human evidence on the question, but it has limits:
- it lasted only two weeks;
- the THC arm was an extract, not a pure isolate;
- it cannot separate what CBD contributed from what the minor compounds contributed.
Newer controlled studies:
- Englund et al. (2023, Neuropsychopharmacology) studied 46 healthy, infrequent cannabis users. They inhaled vaporised cannabis containing 10 mg THC plus 0, 10, 20 or 30 mg CBD. CBD did not significantly change THC’s effects on memory, psychotic-like symptoms, subjective effects or physiological measures. This undercuts the popular claim, which goes back to McPartland & Russo 2001, that CBD “buffers” THC at commonly sold ratios.
- Spindle et al. (2024, Drug and Alcohol Dependence) was a double-blind study in 20 healthy adults. They inhaled vaporised THC with or without d-limonene. Limonene reduced THC-induced anxiety and paranoia, more so at higher doses, and did not change THC’s other effects. This is the first controlled human evidence that a terpene can modulate a THC effect. It is still a single small study that has not been replicated.
We are not aware of any adequately powered trial that has compared all three relevant arms for a common condition: THC isolate, THC + CBD, and full-spectrum extract at the same THC dose. Part of the reason is practical. Such trials need preparations with a defined, reproducible composition, and quality control of research cannabis has often varied widely (Thomas & Pollard, “Preparation and Distribution of Cannabis and Cannabis-Derived Dosage Formulations for Investigational and Therapeutic Use in the United States”, Frontiers in Pharmacology 2016;7:285).
The isolate counterpoint: Epidiolex. Epidiolex is a purified CBD oral solution. It is approved in the US and EU for seizures in Lennox-Gastaut syndrome, Dravet syndrome and tuberous sclerosis complex. Its success does not refute the entourage hypothesis, which never claimed that isolates cannot work. It does show that isolates have real clinical value.
How the Term Became Marketing
- A legitimate finding in a narrow setting: the endocannabinoid entourage effect (1998).
- Extended as a hypothesis to a commercially attractive area: phytocannabinoids plus terpenes (Russo 2011).
- The hypothesis is cited as evidence, and the word “hypothesis” is dropped.
- “Full-spectrum is better than isolate” becomes a sales axiom.
- Specific terpene profiles are marketed as producing specific effects. For example, a high-myrcene cultivar is described as “deeply relaxing” as though data on myrcene’s effects in humans existed.
At each step the claimed certainty went up while the evidence stayed where it was.
Full-Spectrum, Broad-Spectrum, Isolate
Isolates (pure CBD, pure THC) have the cleanest pharmacology: the dose–response is predictable and trials are easy to design. For specific epilepsy syndromes, pharmaceutical CBD isolate is the option with trial evidence; prescribing is a specialist decision.
Broad-spectrum products keep other cannabinoids and terpenes but have the THC removed, usually by distillation or chromatography. If the most important interaction is between THC and CBD, as Johnson 2010 hints, then removing THC also removes that part of any combined effect.
Full-spectrum products keep the cannabinoid and terpene profile of the source plant. They fit Russo’s hypothesis best but are the hardest to standardise. Minor cannabinoids and terpenes vary from batch to batch, which makes dosing harder. Local rules decide what a “full-spectrum” product may legally contain; for Thailand, see the box below.
For adult consumers where legal, and without a specific medical diagnosis, the difference between a well-made full-spectrum product and a well-made broad-spectrum one is unlikely to be noticeable. There are no human trials comparing them. Product quality, accurate labelling and storage probably matter more.
Thailand: what the rules say
- CBD and hemp-extract products are regulated by the Thai FDA. Labels and advertising must stay within approved claims, and an “entourage effect” may not be claimed as a health benefit. In 2026 a Ministerial Regulation B.E. 2569 on cannabis/hemp extracts (in force 26 April 2026) tightened licensing.
- Cannabis flower is a controlled herb under the Notification of the Ministry of Public Health Re: Controlled Herbs (Cannabis) B.E. 2568 (issued 26 June 2025). It may be dispensed only against a prescription on form PT 33 from an authorised practitioner. Each prescription covers no more than 30 days and cannot be refilled, and the flower must come from a GACP-certified cultivation site. Flower may not be sold or advertised online, and recreational use is not permitted.
- The Cannabis Control Act is still a draft. The Cabinet approved it on 22 September 2026 and sent it to the House.
- Epidiolex: we could not confirm that it is registered in Thailand. In September 2026 the Thai FDA was reported to have registered LEPSIA, a pure-CBD anti-seizure medicine for children developed by Chiang Mai University. Use for epilepsy is a specialist decision.
- More detail: forum topics 1293 and 1318.
What the Evidence Supports in 2026
Supported:
- The endocannabinoid entourage effect (Ben-Shabat, Mechoulam et al. 1998), in its original context.
- Nabiximols is effective for MS spasticity. In one cancer-pain trial it beat placebo where a THC extract did not (Johnson 2010).
- β-Caryophyllene binds CB2 receptors (Gertsch 2008).
- Some animal studies show whole-plant extracts behaving differently from isolates (e.g., Gallily 2015).
Plausible, with early or mixed human data:
- Terpenes changing THC effects in humans: one small controlled trial found limonene reduced THC-induced anxiety (Spindle 2024). It has not yet been replicated.
- CBD changing THC effects: not seen at CBD:THC ratios from 1:1 to 3:1 with 10 mg THC (Englund 2023). A pharmacokinetic interaction is possible but probably of little clinical relevance at nabiximols doses.
- Minor cannabinoids (CBG, THCV, CBC) in full-spectrum products meaningfully changing outcomes.
Contested:
- Terpenes acting at CB1/CB2 receptors: two in-vitro studies were negative (Santiago 2019; Finlay 2020) and one mouse/in-vitro study was positive (LaVigne 2021).
Not supported:
- Specific terpene profiles reliably predicting specific effects.
- Full-spectrum products being clinically superior to isolates for all uses.
Conclusion
The entourage effect is a legitimate scientific concept. It began in endocannabinoid biology and was extended to plant compounds as a hypothesis. It now has a small amount of direct human data:
- one three-arm cancer-pain trial that favours THC:CBD over a THC extract;
- one small terpene trial;
- one well-controlled trial in which CBD did not blunt THC.
That is not enough to justify the confident claims made in retail settings.
The honest position is informed uncertainty. Whole-plant cannabis plausibly contains relevant complexity beyond THC and CBD. How large that effect is, and whether it matters clinically in humans, is not established, and predictions of effects from terpene profiles lack an evidence base. Medical patients should discuss product type with a physician who knows cannabis pharmacology. Adequately powered head-to-head trials would settle much of what remains uncertain.