Reading a terpene profile: share, mg/g and totals

A terpene result looks simple: a list of names with numbers beside them. But the same flower can be described three ways, by labs that looked for different lists of compounds, at different points in the flower’s life. This lesson shows how to read terpene numbers without being misled by any of that.

Lesson 6 covers what terpenes are. Here we stay with the numbers.

The panel decides what you can see

A lab doesn’t search for every terpene. It measures a fixed list, its panel, and panels vary a lot. In the six-lab dataset behind Smith et al. (2022), each lab “had a unique set of cannabinoids and terpenes that they measured,” so the authors kept only the 14 common terpenes measured across all labs.¹ Among our own sources, one registry reports about 45 terpenes per sample, while another dataset reports nine.²

So before reading any terpene number, ask which terpenes were on the list. A profile from a short panel can’t show what it didn’t look for.

Three ways to state the same amount

Here are the terpene results from the invented example COA in Lesson 9:

Terpene	mg/g	% w/w	Share of measured total
β-myrcene	6.1	0.61%	39%
β-caryophyllene	3.4	0.34%	22%
Limonene	2.8	0.28%	18%
α-humulene	1.1	0.11%	7%
α-pinene	0.9	0.09%	6%
β-pinene	0.7	0.07%	4%
Linalool	0.6	0.06%	4%
Total	15.6	1.56%	100%

Example data, one invented sample. Shares are rounded.

• mg/g is milligrams of the terpene per gram of sample. It’s an absolute amount.
• % w/w is the same thing divided by 10: 6.1 mg/g is 0.61%.
• Share of the measured total divides each terpene by the panel’s total: 6.1 ÷ 15.6 = 39%. It describes the shape of the profile, not the amount.

The first two answer “how much?” The third answers “in what proportions?” They behave very differently over time, which is the next point.

Not detected, below the limit, not measured

A blank in a terpene list can mean three different things:

1. ND: looked for, and below the method’s detection limit.
2. <LOQ: looked for and seen, but too little to quantify.
3. Not on the panel: never looked for.

None of these is a measured zero. In our pipeline, a terpene that isn’t on a lab’s panel is treated as missing, not zero, and only terpenes on that lab’s panel count toward its share calculation.² Otherwise a lab with a short panel would make every terpene it skipped look absent.

Why totals fall

Terpenes are volatile, and the lighter ones leave first. A 1996 study steam-distilled the volatile oil from fresh cannabis buds and from buds dried and then stored at room temperature in paper bags.³

• The oil content fell by 31.0% after one week of drying, 44.8% after a further month of storage and 55.2% after three months.
• The mix shifted too. Monoterpenes made up 92.5% of the fresh oil and 62.0% after three months, while sesquiterpenes, the heavier class, rose from 6.8% to 35.6%.

That is one study, with older methods and one set of conditions. Still, it shows why a terpene total can say as much about the sample’s handling and age on the day it was tested as about the plant. Two packages of the same flower, tested months apart, can have very different totals.

Regulators treat totals cautiously too. Ohio’s April 2025 guidance prohibits labs from reporting other calculated results, “Total Terpenes” included, on certificates of analysis for now.⁴

How we build a profile

Because shares hold up better than totals, our terpene profiles are built from shares.²

1. A common core. Comparisons between strains use a core set of terpenes that most panels include, so a long panel doesn’t distort a short one.
2. Shares within each sample’s own panel, then combined across samples with the same weights and caps as THC (Lesson 10): no lab or producer above 30% of the weight.
3. A grade for the profile, using the same A–D sample rules as THC.
4. The total is capped at grade C. We show a terpene total in mg/g, but because it depends so much on drying and storage it can never be graded higher than C, however many samples we have.

Reading the bloom chart

Each strain page draws its profile as a bloom.

How to read a bloom

Scroll sideways to see the whole diagram.

When you read one:

• Petal size is share, not amount. A big myrcene petal means myrcene leads the profile, not that the flower holds a lot of it.
• Small terpenes are left off below 5% so the shape stays readable. The full list sits underneath.
• The grade and n apply to the whole profile. A grade D strain shows “Not yet measured” instead of a bloom.

One strain’s terpene profile: Durban Poison

Durban Poison: share of total measured terpenes, 382 lab profiles, grade A · release 2026-10-21

• Terpinolene382 profiles · grade A

  8.8540.3%

• Myrcene382 profiles · grade A

  2.3411.3%

• Pinene382 profiles · grade A

  2.259.3%

• Ocimene172 profiles · grade B

  2.158.5%

• Caryophyllene382 profiles · grade A

  1.36%

• Limonene382 profiles · grade A

  0.974.7%

• Humulene382 profiles · grade A

  0.532.3%

• Linalool382 profiles · grade A

  0.31%

What a profile can’t tell you

A terpene profile is chemistry from a lab panel. It can’t fully predict aroma, because some of the most distinctive notes come from compounds that terpene panels don’t measure (Lesson 7). And it isn’t a guide to anything beyond chemistry.

This chart shows chemistry and aroma from lab tests. How terpenes affect people is not well studied, so we don’t list effects.

Check yourself

Sources

1. Smith CJ, Vergara D, Keegan B, Jikomes N. The phytochemical diversity of commercial Cannabis in the United States. PLOS ONE 17(5): e0267498, 2022.
2. CannabisStrains.com. Methodology and data pipeline (release 2026-10-21, method version 1.0.0).
3. Ross SA, ElSohly MA. The volatile oil composition of fresh and air-dried buds of Cannabis sativa. Journal of Natural Products 59: 49–51, 1996.
4. Ohio Division of Cannabis Control. Updated guidance: potency testing and reporting, April 2025.