The short answer
For legal cannabis in the U.S., potency testing mostly means HPLC: high-performance liquid chromatography. A lab grinds or homogenizes the sample, extracts cannabinoids with solvent, separates them on a column, and measures them with a UV or mass spec detector. GC, gas chromatography, is common for terpenes and residual solvents, and it can run cannabinoids too, but the hot inlet converts THCA to THC. That makes GC a poor fit when you need the acid cannabinoid profile. Near-infrared and other rapid tools are screens, not release methods. The label number depends on sampling, prep, calibration, and the lab's quality system.
Cannabis Product Lab Test Certificate: How to Read a COA
What potency testing measures
Labs report individual cannabinoids: THC, THCA, CBD, CBDA, CBG, CBN, CBC, and THCV are the usual set. Potency labels use total THC, which is THC plus THCA times 0.877. The 0.877 factor accounts for the lost carbon dioxide when THCA converts to THC. Total CBD works the same way. If a COA lists 20% THCA and 1% THC, total THC is about 18.5%. Moisture matters too. A bone-dry flower sample can look hotter by weight than a sample at normal moisture. Potency is a mass fraction, usually percent by weight or mg per gram.
Full Panel Lab Test for Cannabis: What You Need to Know
HPLC: the workhorse
Most accredited labs use reverse-phase HPLC with a C18 column, a water and acetonitrile or methanol gradient, and UV detection around 220 nm or a diode array detector. LC-MS/MS shows up for low-level work in edibles and complex matrices. HPLC does not heat the sample in an inlet, so THCA and THC stay separate. That makes it the better choice for flower, concentrates, tinctures, and products where the acid form matters. The tradeoffs are real: certified reference standards, skilled analysts, and matrix effects in edibles. A gummy or chocolate sample needs a clean extraction and a calibration curve that matches the matrix, or the result can drift.
What Does CBD Lab Tested Mean? Reading a COA Without the Hype
GC: where it helps and where it misleads
GC separates volatile compounds and pairs well with a flame ionization detector or mass spec. For terpenes and residual solvents, it is a strong tool. For cannabinoids, the inlet heat is the problem. At typical inlet temperatures, THCA loses CO2 and becomes THC. A GC potency run on flower can report a total THC number that looks like decarbed potency, not the raw acid content. Some labs use GC for potency anyway and report total THC after conversion. That can be fine for a rough check, but it is not the method I want for a full cannabinoid profile. If a COA lists THCA from a GC method, ask how the lab handled decarboxylation.
Rapid methods: NIR, TLC, and immunoassay
Near-infrared spectroscopy is fast and useful for process monitoring, but it needs a robust calibration against a primary method like HPLC. TLC is cheap and semi-quantitative, good for a quick look, not for a regulatory label. Immunoassays can screen for cannabinoids in some matrices, but they are not a full potency method. None of these replace HPLC for release testing under state rules. They are filters, not final answers.
Sample prep is half the result
Potency starts with the sample. A stemmy top bud and a popcorn bud from the same batch can test differently. Labs need a representative sample, proper homogenization, and a recorded moisture check. Concentrates need dilution by weight, and edibles need full homogenization because a gummy is not uniform. Extraction efficiency, filtration, and dilution errors all show up in the final number. I also check whether the lab is accredited to ISO/IEC 17025 and licensed by the state. That does not guarantee a perfect result, but it tells you the lab has a quality system behind the method.
How to read a potency COA
When I read a COA, I go straight to the method line, the THCA line, and the total THC math. Then I check moisture, sample type, limit of quantification, and test date. A few habits help:
- Confirm the method: HPLC-UV or LC-MS for cannabinoids, GC for terpenes and solvents.
- Check that THCA is reported separately and total THC uses the 0.877 factor.
- Look for moisture content and whether the result is dry-weight or as-received.
- Check the lab's accreditation and state license number.
- Treat a single potency number as an estimate, not a promise.
Lab-to-lab spread is real. State sampling rules and lab methods matter more than a brand's potency claim. A 25% flower label and a 22% flower label can come from the same batch if the sample and lab differ.
Bottom line
HPLC is the standard for cannabinoid potency because it preserves the acid profile. GC is useful for terpenes and solvents, but heat changes the cannabinoid story. Rapid methods screen, and sample prep decides how much the result means. The number on the label is a lab estimate shaped by method, sample, and quality control.