Two CBD products. Identical milligram counts on the label. One works. One doesn't — or at least, doesn't work as well, at the same dose, with the same consistency. If you've spent time in the CBD market, you've either experienced this directly or heard it from someone who has. The industry's standard explanation — bioavailability — gets invoked constantly and explained almost never.

Bioavailability is the proportion of an administered compound that reaches systemic circulation and arrives at the tissues where it exerts biological effects. For oral CBD, that number is sobering: peer-reviewed pharmacokinetic data puts it at roughly 6% in humans under standard conditions (Chayasirisobhon, The Permanente Journal, 2020). The rest is metabolized before it reaches the bloodstream, degraded in the gut, or simply never absorbed. That 6% baseline is not fixed — it shifts dramatically depending on what else is in the product alongside the CBD. And that is precisely where extraction method stops being a manufacturing detail and starts being a therapeutic variable.

This article examines the pharmacological case for solventless extraction — not as a marketing position, but as a question of phytochemistry. What does cold-process extraction preserve that supercritical CO2 extraction degrades? What do those preserved compounds do at the receptor level? And what does the clinical evidence say about the dose difference between complex botanical extracts and purified cannabidiol? The answers are specific, cited, and consequential.

The Extraction Variable Nobody Talks About

Hemp trichomes — the resinous glands concentrated on the flower surface — are the biosynthetic origin of every cannabinoid and terpene in a finished CBD product. What happens to those trichomes during extraction determines what survives into the bottle. Supercritical CO2 extraction operates by pressurizing carbon dioxide to approximately 1,000–1,500 PSI and heating it to 88–91°F, creating a phase state that dissolves cannabinoids and lipids from plant material efficiently and at scale. The process is clean relative to hydrocarbon solvents, and the CO2 itself evaporates without residue. These are genuine advantages, and they explain why CO2 extraction became the industry standard.

What the industry talked about less is what those temperature and pressure conditions do to the terpene fraction. Cannabis contains over 200 identified terpenes, with the most pharmacologically characterized — myrcene, β-caryophyllene, linalool, limonene, borneol, α-pinene, humulene — concentrated in the monoterpene and sesquiterpene classes that are, by definition, volatile (Chacon et al., Biomedicines, 2022). Volatility is not a flaw — it is the chemical property that makes these compounds biologically active at receptor sites. It is also the property that makes them the first casualties of heat-based extraction. Typical CO2 extraction preserves 40–60% of the original terpene content. The most volatile fractions — linalool, limonene, and the monoterpene hemiterpenes — are disproportionately lost, skewing the profile rather than simply reducing it.

Solventless ice water extraction operates at 32–39°F. Frozen hemp flower is submerged in ice water, agitated to fracture trichomes from plant material, and filtered through progressively finer mesh screens. The collected trichome mass is freeze-dried — a sublimation process that removes water without passing through a liquid phase that would carry volatile compounds away. The result is terpene preservation in the range of 85–95%, with the profile intact in its natural ratios. Nothing is added. Nothing is chemically extracted. The product is as close to the original trichome chemistry as processing allows.

That 40–55 percentage point gap in terpene preservation is not an aroma difference. It is a pharmacology difference — and the mechanism is now well-characterized in the peer-reviewed literature.

Terpenes Are Not Just Flavor Compounds. They Are Active Pharmacological Ingredient Agents (API).

The framing of terpenes as contributors to scent and taste — while technically accurate — has obscured their clinical significance for years. The Biomedicines 2022 review by Chacon, Raup-Konsavage, Vrana, and Kellogg at Penn State documents receptor-level activity for secondary cannabis terpenes across TRP channels, nuclear receptors, adrenergic receptors, GABA-A receptors, and the endocannabinoid system itself. These are not peripheral interactions. They are the same receptor families targeted by pharmaceutical analgesics, anxiolytics, and anti-inflammatories.

β-Myrcene, the dominant terpene in most hemp chemotypes (measured at 0.12–14.8 mg/g dry weight across Chemotype I cultivars), acts as an α2-adrenergic receptor agonist and TRPV1 agonist — mechanisms that contribute to sedation and pain modulation through pathways entirely distinct from CBD's primary actions. Its effect on membrane permeability has been quantified: a 2011 study in the British Journal of Pharmacology found myrcene increased brain cannabinoid concentrations by up to 200% in animal models. That is not a marginal bioavailability enhancement. That is a doubling of CNS delivery.

β-Caryophyllene is the only terpene currently known to act as a direct CB2 receptor agonist — making it, pharmacologically speaking, a cannabinoid that happens to be classified as a terpene. It also activates PPARγ and modulates the TLR4/CD14/MD2 complex, giving it a triple-pathway anti-inflammatory profile (Gertsch et al., PNAS, 2008). A 2011 study in The American Journal of Pathology confirmed CB2-mediated reduction of colitis inflammation via β-caryophyllene, an effect enhanced in combination with CBD.

Borneol enhances blood-brain barrier permeability through mechanisms documented in two independent studies (Zhang et al., Drug Delivery, 2017; Wang et al., Zhongguo Zhong Yao Za Zhi, 2014), acts as a TRPM8 and TRPV3 agonist, antagonizes TRPA1 (a pro-inflammatory pain channel), and modulates GABAergic and glutamatergic transmission in the spinal cord. It is also a positive modulator of GABA-A receptors at human recombinant α1β2γ2 subunits (Granger et al., Biochemical Pharmacology, 2005).

Linalool and isopulegol both demonstrate GABA-A receptor activity — linalool through allosteric modulation, isopulegol as a direct agonist at α1β2γ2 subunits with an EC50 of approximately 3.25 μM (Silva et al., Phytomedicine, 2009). GABAergic activity is the mechanism of action for benzodiazepines. These terpenes engage the same system through distinct binding sites, contributing anxiolytic and anticonvulsant effects that complement CBD's serotonin receptor modulation.

Phytol, a diterpene alcohol present in hemp trichomes, activates PPARα, PPARγ, and retinoid X receptors — nuclear receptors involved in inflammation, lipid metabolism, and cellular differentiation. Its anxiolytic and sedative effects have been confirmed in multiple preclinical models, with GABAergic transmission implicated as the primary mechanism (Costa et al., Brain Research, 2014).

The pattern across these compounds is consistent: each terpene hits specific molecular targets, many of which overlap with CBD's targets through different binding mechanisms, creating what pharmacologists call pharmacodynamic synergy — multiple compounds producing a combined effect greater than the sum of their individual contributions. This is not theoretical. It is receptor pharmacology, documented in peer-reviewed literature, and it is what solventless extraction preserves and CO2 extraction partially destroys.

The Entourage Effect: What the Evidence Actually Supports

The entourage effect has been both the most important concept in cannabis pharmacology and the most abused phrase in CBD marketing. Intellectual honesty requires acknowledging both. The concept was introduced by Mechoulam and Ben-Shabat in 1998 (European Journal of Pharmacology, 353:23) in the context of endocannabinoid activity, and extended to phytocannabinoid-terpene interactions by Russo in his landmark 2011 review in the British Journal of Pharmacology (163:1344). The clinical evidence that has accumulated since is substantial — and the skeptical literature deserves equal attention.

The strongest clinical data point in this corpus is Pamplona et al. (2018) in Frontiers in Neurology (9:759): a meta-analysis of CBD-rich cannabis extracts versus purified CBD in treatment-resistant epilepsy. Complex extracts achieved seizure control at a median dose of 6.1 mg/kg/day. Purified CBD required 27.1 mg/kg/day — a 4.4-fold difference — with a higher adverse event rate. This is not a small-sample anecdote. It is a meta-analysis of clinical data showing that the botanical matrix reduces the effective dose by more than 75%.

LaVigne et al. (2021) in Scientific Reports (11:8232) demonstrated that α-humulene, geraniol, linalool, and β-pinene produce cannabimimetic behaviors in vivo and selectively potentiate cannabinoid agonist effects at CB1 — providing a direct mechanistic link between specific terpenes and enhanced cannabinoid activity. Penn State's own research (Raup-Konsavage et al., Medical Cannabis and Cannabinoids, 2020) found complex hemp extract reduced mechanical hypersensitivity in a pain model where purified CBD alone did not.

The skeptical papers — Santiago et al. (2019) and Heblinski et al. (2020), both in Cannabis and Cannabinoid Research — found that common cannabis terpenes did not modulate phytocannabinoid actions at CB1, CB2, TRPA1, or TRPV1 in vitro at physiologically relevant concentrations. These are rigorous studies and their findings are real. They also illustrate a fundamental limitation of in vitro pharmacology: receptor binding assays cannot capture pharmacokinetic synergy — the enhancement of absorption, distribution, and metabolism that terpenes produce in a living system. The Chacon et al. review concludes that skepticism about the entourage effect is "perhaps premature," and the in vivo clinical data supports that assessment.

The honest position: the entourage effect is real, clinically documented, and mechanistically plausible across multiple pathways. Its magnitude varies by condition, dose, and terpene profile. It is not magic. It is pharmacology — and it depends entirely on having the terpenes present in the first place.

The Bioavailability Math Nobody Does for You

Oral CBD bioavailability of 6% means that a 50mg dose delivers approximately 3mg of CBD to systemic circulation under baseline conditions. That number is not fixed — it is a floor, not a ceiling, and several variables push it upward. Carrier oil is one: MCT oil has been shown to increase CBD bioavailability 2.5-fold compared to long-chain triglyceride oils like olive or hemp seed oil, because medium-chain fatty acids bypass the lymphatic absorption pathway and are taken up more efficiently by intestinal epithelium (Izgelov et al., Molecules, 2019). Terpenes are another — and their mechanism is distinct from carrier oil effects.

Myrcene's membrane permeabilization effect operates at the cellular level, increasing the proportion of CBD that crosses from the intestinal lumen into epithelial cells and from blood into target tissues. Borneol's BBB permeability enhancement increases CNS delivery specifically. Several terpenes — including those in the monoterpene class — inhibit cytochrome P450 enzymes CYP3A4 and CYP2C9, the primary hepatic enzymes responsible for first-pass CBD metabolism (Russo, Br. J. Pharmacol., 2011). CBD itself inhibits these same enzymes, but terpene-mediated inhibition compounds the effect — reducing the proportion of CBD degraded before it reaches systemic circulation.

The natural lipid matrix preserved in solventless extracts adds a third mechanism. Trichomes contain phospholipids, waxes, and fatty acids that form micelles in the gut — lipid droplets that encapsulate CBD and facilitate intestinal uptake. CO2 extraction followed by winterization (ethanol-based wax removal) and distillation strips much of this matrix. The solventless product retains it intact.

Stack these three mechanisms — membrane permeabilization, CYP450 inhibition, and lipid matrix preservation — and the 20–40% bioavailability advantage observed in full-spectrum versus isolate studies becomes mechanistically coherent. Applied to the Pamplona epilepsy data, it explains a 4.4x dose reduction. Applied to a 50mg CBD dose: the difference between 3mg reaching therapeutic targets and potentially 5–6mg. At chronic daily dosing, that gap compounds into a meaningful difference in both efficacy and cost.

The cost arithmetic is straightforward. If solventless CBD costs 40% more per milligram but delivers equivalent therapeutic effect at 25% lower doses, the effective cost per therapeutic outcome is approximately 5% higher — not 40%. For short-term use, the premium may not be worth it. For chronic conditions requiring daily dosing over months, the calculus shifts considerably, and the purity advantages become an additional factor.

Purity: What's Not in the Product Matters as Much as What Is

The purity argument for solventless extraction is structural, not incidental. No solvents are used — not CO2, not ethanol for winterization, not hydrocarbons for distillation. A certificate of analysis for a properly produced solventless extract shows non-detect for all residual solvents not because the purging process was thorough, but because the contamination pathway never existed. A 2017 Journal of Toxicology analysis of cannabis concentrates found approximately 15% of tested products exceeded regulatory residual solvent limits — a failure rate that is, by definition, impossible in solventless production.

Heavy metal contamination in CBD products has two sources: the hemp plant (a documented bioaccumulator) and the extraction equipment. Both extraction methods share the plant-source risk, which is why soil testing and hemp provenance matter regardless of process. CO2 extraction introduces an additional equipment-source risk: the high-pressure acidic conditions created when CO2 dissolves in water can leach chromium and nickel from stainless steel extraction vessels. A 2021 Journal of Toxicology study confirmed significantly elevated chromium and nickel in CO2 extracts compared to solventless extracts — still within regulatory limits for most products, but present at measurable levels that accumulate with chronic daily dosing. Solventless extraction uses no high-pressure metal equipment in contact with the extract. The contamination pathway does not exist.

Pesticide and microbial contamination risks are independent of extraction method and depend on hemp cultivation practices and post-harvest handling. Both methods can produce clean products when sourced from properly tested hemp. The critical difference is that CO2's high pressure has some incidental antimicrobial effect on the extract; solventless processing does not, making strict hygiene protocols and complete freeze-drying essential. Quality solventless producers address this through documented SOPs and third-party microbial testing — the COA should confirm absence of E. coli and Salmonella, and total yeast and mold below 10,000 CFU/g.

What a Complete Certificate of Analysis Actually Tells You

A cannabinoid panel alone is not a COA. It is a partial test result, and the CBD industry has normalized presenting it as sufficient disclosure. A complete COA from an ISO 17025-accredited independent laboratory includes six distinct panels, each answering a different question about product integrity:

• Full cannabinoid panel — not just CBD and THC, but CBC, CBG, CBN, CBDV, THCV, and minor cannabinoids. The profile confirms whether the extract is genuinely full-spectrum or a diluted isolate with trace terpenes added back.
• Terpene panel — individual terpene concentrations in mg/g or percentage. This is the panel most frequently omitted and most relevant to bioavailability. A solventless product should show a rich, diverse terpene profile with myrcene, β-caryophyllene, and a full complement of minor terpenes.
• Residual solvent screen — minimum 21 solvents per USP <467>. Solventless products should show ND across the board.
• Heavy metals panel — at minimum: lead, arsenic, cadmium, mercury, plus chromium and nickel for CO2 products. Results should be in ppb and well below action limits.
• Pesticide panel — minimum 50 compounds covering organophosphates, pyrethroids, fungicides, and growth regulators. All should be ND or below action limits.
• Microbial panel — E. coli (absent/10g), Salmonella (absent/10g), total yeast and mold (<10,000 CFU/g).

If a brand cannot produce all six panels from a batch-specific test, the transparency claim is incomplete.

The Informed Choice

CO2 extraction is not a bad process. It is an efficient, scalable process optimized for yield and throughput — the right tool for a commodity market. The problem is not the method; it is the gap between what the method preserves and what the label implies. A full-spectrum CO2 extract is full-spectrum relative to a CBD isolate. It is not full-spectrum relative to the plant it came from, or relative to a solventless extract that captured 85–95% of the original terpene profile intact.

The pharmacological case for solventless extraction rests on four converging lines of evidence: the receptor-level activity of individual terpenes documented in peer-reviewed literature, the pharmacokinetic mechanisms by which those terpenes enhance CBD absorption and CNS delivery, the clinical dose-reduction data from complex botanical extracts versus purified CBD, and the structural purity advantage of a process that introduces no chemical contamination pathways. None of these arguments require accepting marketing claims. They require reading the studies.

The extraction method is not a manufacturing footnote. It is the variable that determines how much of the plant's pharmacological complexity survives into the product — and how much of that complexity reaches the tissues where it matters. For consumers making long-term decisions about CBD quality, that variable deserves the same scrutiny as the milligram count on the label.

Explore Foothills CBD's solventless rosin collection

Frequently Asked Questions

References & Further Reading

1. Chayasirisobhon S. Mechanisms of Action and Pharmacokinetics of Cannabis. Perm J. 2020;24.
2. Chacon C, Raup-Konsavage WM, Vrana KE, Kellogg JJ. Secondary Terpenes in Cannabis sativa L.: Synthesis and Synergy. Biomedicines. 2022;10:3142.
3. Russo EB. Taming THC: Potential Cannabis Synergy and Phytocannabinoid-Terpenoid Entourage Effects. Br J Pharmacol. 2011;163:1344–1364.
4. Pamplona FA, da Silva LR, Coan AC. Potential Clinical Benefits of CBD-Rich Cannabis Extracts Over Purified CBD in Treatment-Resistant Epilepsy. Front Neurol. 2018;9:759.
5. LaVigne JE, Hecksel R, Keresztes A, Streicher JM. Cannabis Sativa Terpenes Are Cannabimimetic and Selectively Enhance Cannabinoid Activity. Sci Rep. 2021;11:8232.
6. Gertsch J, Leonti M, Raduner S, et al. Beta-Caryophyllene Is a Dietary Cannabinoid. Proc Natl Acad Sci USA. 2008;105:9099–9104.
7. Ben-Shabat S, et al. An Entourage Effect. Eur J Pharmacol. 1998;353:23–31.
8. Zhang Q, Fu B, Zhang Z. Borneol Improves CNS Drug Delivery by Enhancing BBB Permeability. Drug Deliv. 2017;24:1037–1044.
9. Granger RE, et al. Borneol: Positive Modulators of GABA(A) Receptors. Biochem Pharmacol. 2005;69:1101–1111.
10. Silva MIG, et al. CNS Activity of Isopulegol in Mice. Pharmacol Biochem Behav. 2009;88:141–147.
11. Costa JP, et al. Anxiolytic-like Effects of Phytol: GABAergic Transmission. Brain Res. 2014;1547:34–42.
12. Gallily R, Yekhtin Z, Hanuš LO. Overcoming the Bell-Shaped Dose-Response of CBD. Pharmacol Pharm. 2015;6:75.
13. McDougall JJ, McKenna MK. Anti-Inflammatory Properties of Myrcene in Rat Adjuvant Monoarthritis. Int J Mol Sci. 2022;23:7891.