The short answer: a sulfur compound carries the skunky note

Weed can smell like skunk because of volatile sulfur compounds, especially 3-methyl-2-butene-1-thiol. Researchers identified this compound as a major contributor to the characteristic skunky note in the cannabis samples they tested. Terpenes contribute to the broader aroma, but a terpene list alone does not explain that note.

“Volatile” means a substance can enter the air. “Sulfur compound” describes its chemistry. Neither term tells you how a product will affect someone. The question here is what reaches the nose, not what causes intoxication.

Sources: Identification of a New Family of Prenylated Volatile Sulfur Compounds in Cannabis (2021)

How did researchers test the explanation?

In a 2021 study, researchers examined 13 cultivars with instruments that separated compounds and detected sulfur. Four people assessed pungency. The investigators also compared a reconstructed aroma with and without the suspected sulfur compound; adding it made the mixture more closely resemble the flower.

That reconstruction is more informative than simply finding a chemical in a fragrant sample. It asks whether changing the mixture changes the smell. Still, a small panel and selected samples do not represent every plant or every nose.

The paper reports no external funding and discloses a patent application related to the findings. Its authors include researchers from a commercial aroma company. Those interests belong beside the finding, not hidden behind a benefits claim.

Sources: Identification of a New Family of Prenylated Volatile Sulfur Compounds in Cannabis (2021)

Compare what people smelled with what a lab measured

A later study paired sensory descriptions with chemical measurements. The explorer below uses its public data. Choose a sample, then compare the two lists. They answer different questions: what people noticed, and how much of selected compounds the laboratory reported.

For example, the number beside “Citrus” counts people who selected that word. The percentage beside limonene describes a chemical measurement. Do not read those numbers as interchangeable scores or assume that matching words establish cause and effect.

Anonymous study codes, not product recommendations. Type I means THC-dominant; type III means CBD-dominant. These are observations of unburned flower.

What people smelled

Most selected descriptions for sample 11. Each number counts people out of 21; it is not smell intensity.

  • Candy13 / 21
  • Herbal10 / 21
  • Fruity9 / 21
  • Woody7 / 21
  • Berry7 / 21

“Skunky” was selected by 0 of 21 people.

What the lab measured

The five largest reported terpene/terpenoid values for this sample, as a percentage of sample weight.

  • (R)-(+)-Limonene1.52%
  • ß-Myrcene0.227%
  • ß-Caryophyllene0.16%
  • (-)-ß-Pinene0.148%
  • α-Pinene0.131%

Sulfur analysis: not performed for this sample group. Missing does not mean zero.

All observations and how to read them

You can select several smells, so counts do not add up to 21. Tied values retain the workbook order. Terpene values are not adjusted for moisture; a reported zero does not prove absolute absence. Sulfur peak-area measurements use a different scale and are not plotted against these percentages.

  • Candy13 / 21
  • Herbal10 / 21
  • Fruity9 / 21
  • Woody7 / 21
  • Berry7 / 21
  • Earthy6 / 21
  • Musty5 / 21
  • Citrus4 / 21
  • Floral4 / 21
  • Tropical4 / 21
  • Vomit/fecal4 / 21
  • Straw3 / 21
  • Chemical3 / 21
  • Black Tea3 / 21
  • Cheesy2 / 21
  • Fuel2 / 21
  • Nutty / Toasted bread2 / 21
  • Cakey2 / 21
  • Creamy2 / 21
  • Pepper1 / 21
  • Doughy/Yeasty1 / 21
  • Ammonia0 / 21
  • Skunky0 / 21
  • Animal0 / 21
  • Baking spice0 / 21
  • (R)-(+)-Limonene1.52%
  • ß-Myrcene0.227%
  • ß-Caryophyllene0.16%
  • (-)-ß-Pinene0.148%
  • α-Pinene0.131%
  • Linalool0.0922%
  • (-)-α-Terpineol0.086%
  • (+)-Fenchol0.0844%
  • Farnesene0.0765%
  • Humulene0.0641%
  • trans-ß-Ocimene0.0549%
  • (±)-trans-Nerolidol0.0466%
  • Terpinolene0.0283%
  • Camphene0.0215%
  • cis-ß-Ocimene0.00718%
  • (-)-Guaiol0%
  • Gamma-Terpinene0%
  • (-)-Caryophyllene Oxide0%
  • α-Bisabolol0%
  • α-Phellandrene0%
  • d-3-Carene0%
  • α-Terpinene0%
  • (+)-Borneol0%

Adapted from Isaacson, Wilson-Poe, Ye, Qian and Shellhammer (2025), supporting data, CC BY 4.0. We joined sample codes, selected the combined 21-person table and rounded displayed chemical values. This comparison does not identify which chemical caused a smell.

Sources: Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma (2025); Supporting data: sensory differentiation of cannabis and hemp aromas (2025)

What the newer aroma study can—and cannot—tell us

The 2025 study evaluated 91 samples. Its combined sensory table uses the 21 people who assessed both sample groups. Participants smelled the flower; they did not smoke or consume it. Terpene profiles alone poorly predicted sensory character in this dataset.

Sulfur testing covered only the CBD-dominant group. Sample handling and testing locations differed between groups, and the authors discuss limits on their statistical models. This is evidence about a selected collection, not a universal aroma calculator.

Vectura Fertin Pharma CH funded the study. The authors declared no competing interests and reported that the funder had no role in the study or publication decisions.

Sources: Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma (2025)

Three ways of examining an aroma

Before interpreting a result, ask which question the method was designed to answer. A number can be precise while answering a narrower question than you expected. These approaches complement one another; none should be mistaken for a complete description of a person’s experience.

Different observations answer different questions
ApproachQuestion it helps answerLimit to keep in view
Terpene panelHow much of the selected terpenes was reported?Does not include every possible odor contributor.
Sulfur-focused analysisWhich sulfur signals did this method detect?A signal is not a sensory rating; methods and units matter.
Sensory assessmentWhich descriptions did people select?A description does not identify its chemical cause.
Aroma reconstructionDoes adding a compound change the reconstructed smell?A result for a tested mixture is not a rule for every sample.

Sources: Identification of a New Family of Prenylated Volatile Sulfur Compounds in Cannabis (2021); Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma (2025)

Does a stronger smell mean stronger effects?

The studies discussed here investigate aroma. They do not validate smell as a way to estimate intoxication, medical benefit or a safe amount to use. Treat a claim about effects as a separate question requiring its own evidence.

Likewise, a pleasant or familiar smell is not a contaminant test. If you are reading a lab report, look at what was actually tested rather than asking the aroma to stand in for missing measurements. Our terpene-profile guide explains the distinction between reported concentrations and the conclusions a reader might be tempted to draw.

Sources: Identification of a New Family of Prenylated Volatile Sulfur Compounds in Cannabis (2021); Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma (2025)

A better way to explore cannabis aroma

Start with the observation: skunky, citrusy, woody or something else. Next, ask what supports the proposed explanation. Is it a laboratory measurement, a panel’s description or an experiment that changed the mixture?

Use those questions to keep curiosity separate from certainty. A useful aroma guide should let you follow the evidence without requiring you to accept a product claim. The chart and compound guides below are places to continue learning; they are not prescriptions for a particular feeling.

  • Check the material: unburned flower and smoke are different questions.
  • Check the measurement: a concentration, a sensory count and a relative instrument signal are different quantities.
  • Check the limits: unavailable data should remain unavailable, not turn into zero.

Sources: Identification of a New Family of Prenylated Volatile Sulfur Compounds in Cannabis (2021); Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma (2025)