The Entourage Effect: What’s Actually Been Shown

Last updated on September 7, 2026 · Originally published February 26, 2017

Reviewed by Nani Frenkel, chief editor

The entourage effect may be the most-cited idea in cannabis – and the least often checked. Ask a dispensary menu and every terpene has a job; ask a skeptic and the whole thing is marketing. The evidence supports neither answer. What it supports is a scoreboard: a handful of specific cannabinoid-terpene pairings that have now been tested properly, some of which produced positive results, some of which did not, and a large remainder that has never been tested at all. This article is that scoreboard. One scoping note: this is the cannabinoid-terpene version of the theory – cannabinoid-cannabinoid interactions, THC with CBD most famously, are their own literature and their own question.

The scoreboard at a glance: limonene + THC for anxiety – promising human evidence, awaiting replication. Pinene + THC for memory – not supported. Terpenes at CB1/CB2 – synergy reported in vitro, one group, unreplicated. Caryophyllene at CB2 – a real mechanism, not an entourage interaction. Myrcene “priming receptors” – unshown. Whole-profile “ensemble” effects – unshown. The details, and why each verdict lands where it does, follow.

Where does the term actually come from?

Not from terpenes. In 1998, Shimon Ben-Shabat, Raphael Mechoulam, and colleagues coined “entourage effect” to describe something happening inside the body: inactive fatty-acid esters that, on their own, showed no cannabinoid-receptor activity, but that amplify the activity of the endogenous cannabinoid 2-AG when present alongside it. The original entourage was endogenous chemistry – a point this article’s earliest version, to its credit, got right when most coverage did not.

The idea jumped to the plant in 2011, when Ethan Russo published “Taming THC” in the British Journal of Pharmacology, proposing that cannabis terpenes might synergize with cannabinoids – pinene for THC’s memory effects, limonene for mood, myrcene for sedation. It is the founding document of the modern entourage conversation, and it is worth being precise about what it was: a review proposing hypotheses, built largely on studies of terpenes alone, often at high doses. Russo did not claim the synergies were proven. The industry that quoted him was less careful.

What did the null studies show?

The first rigorous attempts to test the hypothesis went straight at the most obvious mechanism – do terpenes act at the receptors THC acts at? – and came back empty. The most cited is Santiago and colleagues’ 2019 study, titled, unambiguously, “Absence of Entourage”: six common cannabis terpenoids, tested alone and in combination, neither activated human CB1 or CB2 in their assay nor changed THC’s signal through them. Other groups reported similar results with different methods – Finlay and colleagues in 2020 among them. A 2024 systematic review – which included cannabis-pharmaceutical-company authors and disclosed a formal company-university collaboration – still ended at “remains unproven” for terpene-cannabinoid synergy as a general claim.

So the null camp earned its skepticism. But “the entourage effect is false” is a claim about a whole category, and categories are not how the recent evidence has moved. It has moved one pairing at a time.

The scoreboard

Limonene + THC, for anxiety – promising human evidence, awaiting replication. The strongest entry on the board, stated carefully. In a 2024 double-blind study at Johns Hopkins, Spindle and colleagues – with Russo among the authors – had twenty adults inhale vaporized THC with and without d-limonene. Limonene alone did nothing measurable. At the highest combination, 30 mg THC with 15 mg limonene, ratings of “anxious/nervous” and “paranoid” were significantly lower than with THC alone, while the other measured effects were not significantly altered. That highest-dose condition has important limits: only twelve participants completed it, and it was always administered last rather than randomized, so order and tolerance effects cannot be excluded – the paper itself says so. The dose and ratio were also unlikely to occur naturally in flower, and in the interest of the scrutiny applied elsewhere on this board, the paper’s disclosures include a related patent application and an author’s advisory role with a terpene company. This is a genuine human signal – the first of its kind – but not yet a settled result. Our piece on limonene’s effects on anxiety and depression covers the details.

α-Pinene + THC, for memory – not supported. The same group ran a closely matched design on the pairing Russo’s review made famous: pinene, the acetylcholinesterase story, the hope that a pine-scented cultivar protects your short-term memory. In the 2025 study, pinene did not attenuate THC’s memory or cognitive impairments. Same lab, same rigor, Russo again among the authors – which is exactly why the null deserves as much weight as the limonene success. One pairing produced a positive result; its sibling did not. Our pinene profile has the compound’s fuller story.

Terpenes + THC at CB1 and CB2 – synergy reported in vitro, not independently replicated. In 2023, Raz and colleagues reported in Biochemical Pharmacology that all sixteen terpenes they tested produced some CB1 signalling on their own – at roughly 10–50% of THC’s activation – and that several increased THC-evoked activation, sometimes beyond the sum of the two individual responses. Follow-up work extended terpene activation to CB2, and a 2026 study using isobolographic analysis reported additive or synergistic interactions for several individual terpenes and terpene mixtures at CB1 or CB2 – all of it, so far, in Xenopus oocytes. In mice, a University of Arizona group reported in 2021 that several terpenes produced cannabinoid-like behaviours and additively increased some effects of WIN55,212-2, a synthetic cannabinoid agonist – with the terpenes injected at doses reaching 200 mg/kg, so this demonstrates a possible pharmacological interaction under experimental conditions, not what inhaled flower does; the study is covered in detail in our study report. These findings genuinely conflict with the earlier functional studies, which found neither terpene activation nor modulation of THC signalling. The experiments used different cell systems, receptor-expression conditions, and analytical approaches. Those methodological differences may help explain the disagreement, but they do not resolve it. The positive Xenopus-oocyte findings all come from the same research group and the same heterologous expression system; some of the authors disclosed commercial cannabis affiliations. They provide a credible mechanism to investigate, not evidence that these interactions shape the human cannabis experience.

β-Caryophyllene at CB2 – a demonstrated terpene mechanism, not evidence of an entourage interaction. The cleanest receptor result on the board is not a pairing at all: Gertsch and colleagues showed in 2008 (PNAS 105(26):9099-9104) that β-caryophyllene binds CB2 directly at nanomolar affinity and reduces inflammation in normal mice but not in CB2-knockout mice – the knockout being the control that turns a correlation into a mechanism. That makes BCP a well-established dietary cannabinoid acting on its own, as our caryophyllene profile details – a terpene with genuine cannabinoid-receptor pharmacology, which is not the same as a demonstrated interaction with THC. It stays on the board because it anchors the mechanistic end of the spectrum: what “shown” actually looks like.

Myrcene “helps receptors accept more THC” – unshown. This claim appeared in the original version of this very article, and it circulates everywhere: myrcene primes CB1, opens the blood-brain barrier, turns any cultivar into an indica. No controlled study supports it, and the 0.5% myrcene indica/sativa rule that rode along with it dissolves on contact with actual chemovar data – a correction our myrcene profile makes at length. It stays on the scoreboard as the cautionary row: the most repeated entourage claim is one of the least tested.

What remains unshown?

Most of the evidence above concerns defined combinations – one terpene, one cannabinoid, one outcome. The claim on the product label is grander: that a whole profile of dozens of compounds produces an ensemble effect greater than its parts. No controlled human study has isolated that whole-profile synergy while holding cannabinoid dose and other variables constant. The best evidence usually cited for it comes from the CBD side: in one 2015 mouse inflammatory model, one CBD-rich extract continued producing a dose-dependent response while purified CBD followed a bell-shaped curve. And a 2018 observational meta-analysis of 670 epilepsy patients reported some improvement in 71% of those receiving CBD-rich extracts versus 46% receiving purified CBD, despite lower average CBD doses. But on the stricter and clinically standard endpoint – at least a 50% reduction in seizures – there was no significant difference: 37% with extracts and 42% with purified CBD. The studies were heterogeneous and uncontrolled, so they cannot establish either an extract advantage or synergy – and neither line of evidence can say which co-compounds matter, with minor cannabinoids at least as plausible as terpenes. That is how easily “full spectrum works better” gets constructed from selective numbers – a question with its own CBD-side literature, explored in our piece on terpenes and CBD.

Researchers are beginning to approach the ensemble question more systematically by fractionating extracts and reconstructing the combinations responsible for an observed effect – an approach most prominently associated with Dedi Meiri’s laboratory at the Technion. Some cell studies have identified multi-cannabinoid combinations that reproduce effects of a whole extract. That is a promising experimental method, but so far it has produced evidence mainly for cannabinoid-cannabinoid combinations, not terpene entourage effects – and none of these cancer findings establishes a treatment in patients.

None of this proves the ensemble idea false. It means the ensemble effect should not be sold – or recommended to patients – as an established fact.

What about user reports?

The obvious objection to all of the above is lived experience: millions of people report that different cultivars feel different, and observational datasets find associations between named products or measured chemical profiles and reported effects. That evidence is real and worth collecting – but it cannot, by itself, separate pharmacology from expectation. Many app datasets rely on commercial product names rather than independent chemical analysis; the reports are unblinded, products differ in far more than one compound, and consumers know what they bought. Our own coverage of the aroma research shows how tangled this gets: pleasant aroma, not THC content, best predicted how much people enjoyed a flower. That shows that smell can shape the rated experience; it does not demonstrate cannabinoid-receptor synergy. User reports are where hypotheses for the scoreboard come from. They are not how rows get onto it.

So is the entourage effect real?

The honest answer is that “the entourage effect” is no longer a useful unit of belief. One human trial offers promising but unreplicated evidence of limonene blunting THC’s anxiety at high doses. A closely matched trial did not support pinene protecting memory. One terpene has genuine CB2 pharmacology of its own; sixteen signal at CB1 in oocytes, at fractions of THC’s strength, while other assays find nothing – a conflict the field has yet to resolve. The right question was never “is it real?” but “which pairing, at what dose, shown how?” – and the field has finally begun answering it one pairing at a time. As we said in this article’s first version, nine years and a research literature ago: as the individual components get studied properly, it will change how we understand the plant. It has. Just not in the direction either camp predicted. Learn the compounds themselves in our terpene education hub, start with cannabinoids and terpenes: what’s the difference if the two families need untangling first, and follow the scoreboard here as it fills in – we will keep it current.

References

1. Ben-Shabat S, Fride E, Sheskin T, et al. An entourage effect: inactive endogenous fatty acid glycerol esters enhance 2-arachidonoyl-glycerol cannabinoid activity. European Journal of Pharmacology. 1998;353(1):23-31. https://www.sciencedirect.com/science/article/pii/S0014299998003926
2. Russo EB. Taming THC: potential cannabis synergy and phytocannabinoid-terpenoid entourage effects. British Journal of Pharmacology. 2011;163(7):1344-1364. doi:10.1111/j.1476-5381.2011.01238.x
3. Santiago M, Sachdev S, Arnold JC, McGregor IS, Connor M. Absence of Entourage: Terpenoids Commonly Found in Cannabis sativa Do Not Modulate the Functional Activity of Δ9-THC at Human CB1 and CB2 Receptors. Cannabis and Cannabinoid Research. 2019;4(3):165-176. doi:10.1089/can.2019.0016
4. Spindle TR, Zamarripa CA, Russo E, et al. Vaporized D-Limonene Selectively Mitigates the Acute Anxiogenic Effects of Δ9-Tetrahydrocannabinol in Healthy Adults Who Intermittently Use Cannabis. Drug and Alcohol Dependence. 2024;257:111267. doi:10.1016/j.drugalcdep.2024.111267
5. Kumar L, Spindle TR, Zamarripa CA, Elder HJ, Russo EB, Bigelow G, Vandrey R. The Individual and Interactive Effects of Alpha-Pinene and Delta-9-Tetrahydrocannabinol in Healthy Adults. Medical Cannabis and Cannabinoids. 2025;8(1):144-157. doi:10.1159/000547014. https://karger.com/mca/article/8/1/144/929456/
6. Raz N, Eyal AM, Berneman Zeitouni D, Hen-Shoval D, Davidson EM, Danieli A, Tauber M, Ben-Chaim Y. Selected cannabis terpenes synergize with THC to produce increased CB1 receptor activation. Biochemical Pharmacology. 2023;212:115548. doi:10.1016/j.bcp.2023.115548
7. Raz N, Eyal AM, Fahoum-Khalefa N, Tauber M, Ben-Chaim Y. Selective activation of cannabinoid receptors by cannabis terpenes. Biochemical Pharmacology. 2026;243(Pt 1):117498. doi:10.1016/j.bcp.2025.117498
8. Raz N, Eyal AM, Fahoum-Khalefa N, Tauber M, Schurr N, Ben-Chaim Y. Synergistic and additive terpene-THC interactions in cannabinoid CB1 and CB2 receptors. Biochemical Pharmacology. 2026;251(Pt 2):118185. doi:10.1016/j.bcp.2026.118185
9. Gertsch J, et al. Beta-caryophyllene is a dietary cannabinoid. PNAS. 2008;105(26):9099-9104. doi:10.1073/pnas.0803601105
10. The entourage effect in Cannabis Medicinal products: a comprehensive review. Pharmaceuticals. 2024;17(11):1543. https://research.ulusofona.pt/en/publications/the-entourage-effect-in-cannabis-medicinal-products-a-comprehensi/
11. Overcoming the Bell-Shaped Dose-Response of Cannabidiol by Using Cannabis Extract Enriched in Cannabidiol. Pharmacology & Pharmacy. 2015;6:75-85. https://www.scirp.org/html/5-2500582_53912.htm
12. Pamplona FA, da Silva LR, Coan AC. Potential Clinical Benefits of CBD-Rich Cannabis Extracts Over Purified CBD in Treatment-Resistant Epilepsy: Observational Data Meta-analysis. Frontiers in Neurology. 2018;9:759. doi:10.3389/fneur.2018.00759 (corrigendum: Front Neurol. 2019;9:1050)
13. LaVigne JE, Hecksel R, Keresztes A, Streicher JM. Cannabis sativa terpenes are cannabimimetic and selectively enhance cannabinoid activity. Scientific Reports. 2021;11(1):8232. doi:10.1038/s41598-021-87740-8
14. Finlay DB, Sircombe KJ, Nimick M, Jones C, Glass M. Terpenoids From Cannabis Do Not Mediate an Entourage Effect by Acting at Cannabinoid Receptors. Frontiers in Pharmacology. 2020;11:359. doi:10.3389/fphar.2020.00359