A terpene number tells you how much of one measured group of compounds is in a product. It can't tell you the whole smell. In a 2023 study of 31 ice hash rosins, the compounds that best tracked the sweet, fruity or savory character people noticed were mostly not terpenes at all, and they were present in tiny amounts.

What the study did

Iwan Oswald and colleagues published the work in ACS Omega as "Minor, Nonterpenoid Volatile Compounds Drive the Aroma Differences of Exotic Cannabis". They bought ice hash rosin from dispensaries around Los Angeles, chose samples to cover a wide range of aromas, and gave priority to popular ones. Two-dimensional gas chromatography mapped the volatile compounds in each sample. Separately, seven cannabis users, deliberately untrained, smelled each sample at room temperature, described it in their own words, and scored it from 0 to 100 for sweetness or fruitiness. The authors call that score "exotic."

What they found

By concentration, terpenes typically made up more than 95 percent of the volatile compounds measured. Yet total volatile concentration had essentially no relationship with the exotic score (r = 0.03): samples with more volatiles didn't score as sweeter or fruitier. That total covers everything volatile the study quantified, so it isn't the same figure as a total-terpene percentage on a COA, which the study didn't test directly.

The sharpest example is GMO, which scored at the bottom, and Grape Pie × Do-Si-Do, which scored at the top. Their monoterpene profiles correlated at r = 0.91. Limonene and myrcene, two of the terpenes most often named on menus, showed little relationship to the score either.

What did track the differences were minor compounds the authors call flavorants, often less than 0.05 percent of a sample's mass:

  • A family of sulfur compounds built around a 3-mercaptohexyl group, also found in passionfruit and grapefruit, went with strong citrus and tropical notes.
  • Skatole, a pungent indole compound, went with savory and "chemical" samples.
  • Esters, more than 30 of them, went with fruit.

Context mattered. In one sample, Garlic Cocktail #7, the tropical sulfur compounds appear to have masked the skatole. The "gas" descriptor came up for every sample except Trainwreck; the authors trace it to a different family of sulfur compounds described in their earlier work.

Their conclusion for labels is direct: COAs and packaging that list only terpenes may lead people to treat terpenes as the source of a product's aroma.

How far it reaches

This is 31 rosins from one market, picked for variety rather than at random, and judged by a small panel. The authors flag a likely mislabeled sample and had only one terpinolene-dominant sample. The findings are correlations within that set. The research was financially supported by Abstrax Tech Inc., and several authors disclose patent filings related to the results.

Reading a terpene list after this

The list still has uses. It names the dominant terpenes, and the authors hypothesize that terpene profiles may still shape broad aroma groupings. Just treat it as one part of the description, the large-volume part. When a label lists terpenoids, California's sampling checklist for distributors has them check those figures against the COA, which DCC's testing page says covers cannabinoids and terpenes. The required categories don't name trace sulfur compounds or esters.

The gap is easy to see in reviewer language. The Nasha Poochies Papaya review reports that cracking the jar let "the papaya fruitiness" take over the room. The PUA Diesel review, of a bubble hash rather than a rosin, describes the flavor as "like sucking on a tractor exhaust pipe." Resin Manual has no chemical analysis for either jar, and nothing here identifies what produced those impressions. They're the kind of description a terpene percentage was never going to reproduce.

For what happens to aroma over processing and storage, see what terpene preservation means. For the rest of the package, how to read rosin labels, and for weighing smell against everything else, how to tell if rosin is good.