Terpenes in Edibles: The Lab Numbers, the Liver, and What 'Live Resin' on a Gummy Actually Buys

Terpenes are real molecules with real pharmacology, and almost none of that pharmacology has been shown in a person who swallowed a gummy. That gap is the honest reading of "live resin," "strain-specific" and "botanical terpenes" on an edible label: they buy a flavor and a story, and a patient paying for them deserves to know which is which. The best human result pairing a terpene with THC, limonene easing anxiety in a Johns Hopkins trial, was vaporized, at 15mg, about 15 times what the authors estimated an average joint holds. Two New York gummies with published lab certificates carry 2.2mg and 2.9mg of terpenes, every compound added together.

The brands do not agree with each other either. Kiva's Lost Farm page promises "a true-to-flower high." Wyld's hemp blog says the plant's native terpenes do not make it through cooking and that botanical ones go in afterward, while its FAQ credits those added terpenes with helping produce calm or body relief. None of those pages prints a terpene number. Lab certificates do, so that is where this starts.

How much terpene is actually in an edible?

About 2 to 3mg in two of the three Snowbelt Cannabis gummies we checked, every compound combined. The New York lab certificates show total terpenes of 0.0622 and 0.0813 percent by weight, which is 2.2 and 2.9mg a gummy; the third reads 27mg. We found no peer-reviewed measurement of terpenes in a commercial edible.

That last sentence is a dated finding: a November 7, 2026 PubMed search of titles and abstracts (terpene, terpenoid, limonene, myrcene, linalool or caryophyllene, with edible, gummy, brownie or confection, with cannabis or marijuana) returned eight papers, and no abstract reports a terpene figure for a product on a shelf. So the numbers come from certificates, and the arithmetic is ours. Snowbelt's Concord Cornucopia report lists 9.07mg of THC per gummy at 2.52mg per gram, so the gummy weighs 3.6g, and 0.0622 percent of 3,600mg is 2.2mg. The table runs that sum on the four edible certificates with a terpene panel we found and were able to open, from two New York producers. That is a spot check, and it says nothing about other states.

Out Cold is the outlier. Snowbelt's page names farnesene as its dominant terpene without saying where it comes from, and farnesene appears in none of the human trials below. The gummy named for lemons had no reportable limonene, which takes some doing. For scale, Smith's 2022 analysis of commercial flower from six states found total terpenes averaging 2 percent by weight, so a 1g joint holds about 20mg before any of it burns, and Spindle's group estimated about 1mg of limonene in that joint. The two low-reading gummies carry seven to nine times less than the joint. (The joint also carries far more THC, so this compares absolute doses, which is how the trials dosed them.)

Parts of the market's own guidance sit a long way from these figures. Mood's buying guide says a live resin gummy's certificate should show terpenes at 3 to 5 percent of finished weight, and a Mellow Fellow guide agrees. Lost Farm's Ohio page lists a 3.6g gummy, so that yardstick means 108 to 180mg of terpenes beside 10mg of THC, an essential oil with a THC problem. Snowbelt's Peach OG rosin, the concentrate itself, tests at 4.93 percent, so the bar asks a gummy to match the jar. TribeTokes says its CBD live resin gummy brings terpenes that isolate gummies leave behind, and the 12-page certificate for its CBN live resin gummy has no terpene line for the gummy.

Do terpenes survive the trip into a gummy?

Partly, and the brands disagree about how much. Wyld's hemp blog says the plant's native terpenes do not make it through cooking and botanical ones go in afterward, while Kiva's Lost Farm line is built on the opposite pitch. Snowbelt's certificates suggest a lab's count in a gummy can exceed what the rosin alone supplies.

The chemistry favors Wyld's candor. A Turin group recovered about 58 percent of a flower's terpenes, mostly monoterpenes, in vapor stripped off under mild vacuum during a 30-minute microwave decarboxylation at 120°C. Now run the ratio. Peach OG rosin holds 4.93 percent terpenes against 68.66 percent total THC, so 10mg of THC from that rosin brings 0.72mg of terpene along. The Concord gummy carries 2.2mg, about three times that, and its largest line is bisabolol where the rosin's is caryophyllene.

Those are different cultivars, so this is no before-and-after. Something besides rosin appears to be in the count, and Concord's product page names a candidate in its grape concentrate. Kiva's Ohio page for a strawberry lemonade gummy, infused with a resin named for lemons, lists exactly one named flavoring, and it is raspberry.

What happens to a terpene after you swallow it?

The liver gets it first, the same first pass that reshapes swallowed THC. In Crowell's 1994 study, seven volunteers ate limonene in a custard, and four hours later the major plasma peaks were its metabolites, with limonene itself a minor component. Oral caryophyllene had no published human blood data until 2022, by its authors' account.

The custard held 100mg per kilo, about 7g for a 70kg adult. Our 11-hydroxy-THC page covers what that route does to THC, and the liver makes no exception for something that smells nice.

Caryophyllene has the cleanest claim on a cannabinoid receptor: Gertsch's 2008 PNAS paper measured CB2 binding at 155 nanomolar. Mödinger's 2022 trial then gave 24 healthy adults 100mg in neat oil and measured a peak of 58 ng/mL, about 285 nanomolar by our conversion, under twice the binding figure after a 100mg dose. The Concord gummy's caryophyllene works out to 0.3mg. Scale the blood level in a straight line, which is crude, and you get about 1 nanomolar, some 170 times below the binding figure.

Swallowed terpenes are also ordinary food. Miller's 2010 pilot gave healthy adults 40oz of limonene-rich lemonade a day for four weeks, 500 to 600mg of limonene daily, roughly 300 times the limonene in a Peach Passion gummy, and it built up in their fat. Nobody has yet sold that lemonade as strain-specific.

Is the entourage effect real?

For terpenes acting at cannabinoid receptors, the evidence mostly says no. The term comes from a 1998 paper on the body's own lipids, and Russo's 2011 review extended it to cannabis terpenes while noting the combined effect still had to be proven. Of the five cell studies we read, three found no terpene effect and two reported one.

Santiago 2019 tested six common terpenes at human CB1 and CB2 and found THC's response unchanged (59 percent without them, 55 with, at CB1, the receptor our endocannabinoid page ties to the high). Finlay 2020 found nothing at either receptor beyond a weak caryophyllene interaction at CB2, and Heblinski 2020 drew the same blank at TRPA1 and TRPV1. The dissent is real too. LaVigne 2021 reported terpenes activating CB1 and producing cannabinoid-like behavior in mice, at 200mg per kilo injected into the abdomen and tested alongside a synthetic cannabinoid rather than THC. Raz 2026 reported stronger-than-additive effects with THC at both receptors in frog eggs, and three of its six authors work for a medical cannabis manufacturer. Those labs disagree about what terpenes do in a dish. None of them fed anyone a gummy.

The closest thing to the comparison a buyer wants is 24 years old. Wachtel's group gave 12 people brownies holding pure THC or whole-plant cannabis at matched THC doses and found similar subjective effects with only minor differences, and 13 smokers in a parallel study answered the same way. Cogan 2020 calls the entourage claims poorly defined and unproven, Simei 2024 found limited evidence of a stable, predictable effect, and the Johns Hopkins group behind the best terpene trials wrote in 2026 that human support remains very limited.

Which terpene claims have a human trial behind them?

Five compounds and two blends have human data worth naming. The trials testing THC with and without a terpene, limonene and pinene, were both vaporized. The swallowed studies either used no THC or, in one softgel with 90mg of terpenes, had no terpene-free arm. The table lists route and dose, as of our November 7, 2026 searches.

Limonene is the real result, and it deserves both its credit and its footnotes. The effect appeared at 15mg, in the 12 of 20 people who did that optional session; 1mg and 5mg, the doses chosen to match a joint, did not differ significantly from THC alone. The authors say oral formulations still need testing, Johns Hopkins has filed a patent on the idea, and co-author Russo advises True Terpenes, which supplied the limonene. Reaching 15mg through Peach Passion gummies at 1.8mg each takes eight or nine of them, about 42mg of THC, which our dosing guide files under very high, by a route nobody tested.

Myrcene earns a flag from the other direction: a capsule worsened simulated driving in a small pilot, while both low-reading gummies report it under a quarter of a milligram. Lavender is the fair counterexample to any claim that swallowed terpenes do nothing. Silexan's capsule, 36.8 percent linalool, beat placebo in a meta-analysis with an employee of drugmaker Schwabe among its authors, at about 29mg of linalool a day, more than 100 times the 0.23mg limit those two gummies report it under. Wang's capsule is the one swallowed trial we found with gummy-sized terpene doses, and with 300mg of CBD in every capsule and authors holding equity in the funder, its sleep gain cannot be pinned on the terpenes. Hausenblas's softgel has the same design problem, at 90mg.

Where do terpenes actually matter in an edible?

In the mouth. Flavor is the job terpenes demonstrably do in a gummy, and palatability counts when a patient has to keep swallowing the thing nightly. The pharmacology lives on other routes: inhaled human trials, a little topical work and a large animal literature, where Russo's strongest examples are mice breathing terpenes.

Across 450 California retailers, Pechmann found flavor-named products took 95 percent of edible revenue in 2020 and 2021, and Sommano's 2020 review ties terpenes to consumer experience and market price. Even the aroma story has shrunk. Oswald's 2023 paper, whose authors have filed patents on the work, traced the sweet, savory and citrus scents of exotic cultivars to minor non-terpene compounds and concluded that terpenes' role in aroma has been overstated.

Russo's mice breathed terpenes from room air and changed their activity at single-digit ng/mL blood levels. Topical work is thin: Miller's 2012 orange-oil massage study was safe in women and did not move their plasma limonene, though mice showed tissue uptake.

Readers have reached both conclusions without a lab. On the Straight Dope boards in 2021, one poster told an anxious user that terpenes decide the character of the high and named myrcene and limonene. Two years later another poster there reported edibles labeled indica, sativa and hybrid all feeling identical, and filed the live resin claim under things hard to separate from lore. Wachtel's brownies side with the second poster.

What are you buying with "live resin" or "strain-specific" on a gummy?

A flavor, a production story and a price difference, with no trial we found showing a different clinical result from a swallowed dose. On one San Rafael menu on November 3, 2026, per our gummies ranking, Lost Farm live resin chews ran $23.50 per 100mg against $17 for Wyld Huckleberry: 65 cents more per 10mg piece.

That is $19.50 over 30 nights, a reasonable price for a gummy you like more and a poor one for an anxiety plan. Wyld's own blog makes the point better than we can: the added terpene mix is how it defines a gummy's sativa, indica or hybrid profile, which makes the word on the tin a recipe decision.

A certificate with a terpene panel takes a minute to read this way, once a brand hands one over. Our Wyld and Kiva comparison found Kiva routing requests through email and Wyld's lookup self-serve only in Arizona and New York. Against whatever number comes back, the figure to hold up is 15mg, the lowest dose of a single terpene on this page with a measured effect in a person.