What gives cannabis its unique aroma? A formal analysis

The aroma of cannabis and hemp flower is one of its most distinguishing characteristics, and is often an important factor in determining patient and consumer preferences. It may also help determine the presence of potentially beneficial active ingredients such as terpenes and volatile sulfur compounds.

This study formally evaluated cannabis flower to determine if its perceived aromatic characteristics correlated strongly with specific molecules, namely terpenes and volatile sulfur compounds. It was published in October 2025 in the journal PLoS One. They did not consider the effects of the molecules on people, only their contribution to the aromatic characteristics of cannabis and hemp flower (Type I and Type III cannabis, respectively).

A trained human panel evaluated 91 different samples of cannabis and hemp flower using a reference standard of 25 distinct aroma descriptors, or adjectives. Most of the panelists had prior experience in food and beverage sensory evaluation, and all attended training sessions covering the aroma reference standards and methodology. They were asked to "Check All That Apply" when selecting among the 25 adjectives to describe the aroma profile of the samples. Samples were not smoked or vaporized - only the aroma emanating from dried and cured cannabis flower was evaluated.

The sensory descriptors used were developed starting with open-ended responses from judges at a 2020 cannabis cultivation event that included 81 unique descriptors. The list was refined to include 25 of the most frequent words. They included fruity, citrus, earthy, floral, skunky, fuel, ammonia, and even vomit/fecal. All descriptors except vomit/fecal, fuel, and chemical had reference standards. The detailed list and reference standards used are in the full-text paper (link below).

For some reason the descriptor "Pine" was not used. "Woody" was meant to include all woody aromas, including "Pine". The authors acknowledged that the "Woody" descriptor was insufficient and should be expanded to include "Pine" and a few other unique "Woody" aromas in the future.

Each panelist evaluation session covered about 10 samples and lasted  30-60 minutes. The 25 terms on the ballot were presented in a randomized order. Panelists could take as much time as they needed to evaluate each sample. There was a one minute break between each sample. 

A quick summary of their findings:

Terpenes

• Terpene profiles alone poorly predicted sensory character
• Total terpene content did NOT correlate with the perceived intensity of the aroma
• Myrcene, Terpinolene, and Limonene accounted for about 80% of the variation
• Terpinolene was the only compound consistently associated with a distinguishable aroma, specifically "citrus" and "chemical". 
• High Limonene content was surprisingly NOT associated with a "citrus" aroma

 

VSCs (Volatile Sulfur Compounds)

• VSCs were also not associated with specific aromas
• Most VSCs could not be identified, however the lab results showed that only 4 accounted for most of the variation
• The VSC content and profiles were fairly consistent across all types of cannabis and had MUCH lower variation than was observed among the terpene profiles
  
• Total VSC content per sample was less than 100 PPM, or less than 0.01% of the total weight of the flower sample

 

Other compounds

• They did not test for anything other than terpenes and volatile sulfur compounds
• Nonetheless, since terpenes and VSCs were poorly correlated with distinct aromas, the authors stated that esters, aldehydes, thiols, and other volatile organic molecules were likely important in the perception of cannabis' aroma

 

Type I versus Type III cannabis:

• Type I (high THC, low CBD) was more frequently described as skunky, musty, and animalic
• Type III (low THC, high CBD) had higher frequencies of citrus, fruity, and candy-like aromas.

 

In summary, much of the common wisdom about the aroma in cannabis flower is not confirmed. There seem to be MANY other compounds, beyond terpenes and volatile sulfur compounds, that contribute to cannabis' aroma.

This observation highlights the fact that cannabis flower contains a LOT of compounds that we don't measure or track. If the average flower has 20% cannabinoids, 4% terpenes, and 0.5% each of volatile sulfur compounds and flavonoids (we're being generous), that leaves a whopping 75% of the mass still unidentified. We know there are esters, aldehydes, and many other compounds that surely contribute to its aroma, but more importantly for patients, they could contribute to its effects. We really owe it to patients to start tracking the "Other 75%".

 

Conclusions in the authors' words:

Few terpenes showed consistent or robust predictive relationships with sensory qualities

Terpinolene emerged as the most salient compound, demonstrating a notable association with both "citrus" and "chemical" aroma attributes.

neither terpene nor volatile sulfur compound profiles strongly predicted sensory perception.

The scattered correspondence between chemical composition and sensory quality suggests that the key perceptual drivers may lie in other, unmeasured compound classes, such as esters, aldehydes, or other volatiles that behave in complex synergistic manners creating characteristic impacts beyond individual terpene contributions.

 

The full text article is here at PubMed Central.

Source:

Isaacson SE, Wilson-Poe AR, Ye T, Qian YL, Shellhammer TH. Beyond potency: A proposed lexicon for sensory differentiation of Cannabis sativa L. aroma. PLoS One. 2025 Oct 21;20(10):e0335125. doi: 10.1371/journal.pone.0335125. PMID: 41118398; PMCID: PMC12539713.