Cannabinoid and Terpene Testing Methods: HPLC, GC and What They Measure

Why testing methods matter

A cannabis lab report is only as meaningful as the method behind it. Two instruments can look at the same sample and report different things, and two labs using the same instrument can still disagree. This page explains the main methods in plain language so you can read a report with the right questions in mind. It describes how measurement works; it is not a guide to making or using any product. For reading the report itself, see How to Read a Cannabis Certificate of Analysis (COA).

What labs are asked to measure

Rules differ by place. As one example read on 2026-10-22, California's Department of Cannabis Control says testing labs check cannabis goods for cannabinoids and terpenes, residual solvents and processing chemicals, residual pesticides, heavy metals, microbial impurities, mycotoxins, moisture content and water activity, and foreign material. [1] This page covers the first group, cannabinoids and terpenes. Contaminant testing has its own pages, linked at the end. Laws change and this is not legal advice.

Liquid chromatography (HPLC)

High-performance liquid chromatography, or HPLC, separates the compounds in a liquid extract of the sample and then measures each one with a detector. A narrative review of quantification methods named gas chromatography (GC) and HPLC as the two most common techniques. [2] Its main advantage is that it can measure both the acidic and the neutral forms of cannabinoids without extra chemical treatment (called derivatisation). [2]

That matters because the plant makes acid forms such as THCA first, and the acid and neutral forms are different compounds (see Cannabis Chemistry: Cannabinoids, Acids and Terpenes Explained). A method that sees both lets a report list them separately. The same review's authors concluded that HPLC paired with tandem mass spectrometry (HPLC-MS/MS) is the ideal method for cannabinoids. [2] That is the reviewers' conclusion, not a rule every laboratory follows.

Gas chromatography (GC)

Gas chromatography heats a sample so its compounds turn to vapor and travel through a column. It is often paired with mass spectrometry (MS) or a flame ionization detector (FID). [2] The same review reports that GC's major advantage is measuring terpenes, but that evaluating acidic cannabinoids with GC requires derivatisation. [2]

How terpenes are measured

Terpenes are the volatile aroma compounds. One approach is headspace GC/MS, which samples the vapor above the material. In a 2023 study, researchers used it on leaves and flowers of hybrid cannabis and detected 26 terpenes, six of them major. [3] Each sample had a different terpene composition. [3]

Terpene numbers are harder to compare across studies than many people expect. A 2023 methods paper noted that literature reports vary widely on the major volatile constituents and the concentrations of individual terpenes, which is why its authors set out detailed procedures for reliable analysis. [4] See also Terpene Testing: How Profiles Are Measured and Reported.

Why two labs can disagree

Methods are only part of the story. An analysis of Washington State's seed-to-sale data found systematic differences in the cannabinoid content reported by different laboratories, and the differences persisted after controlling for plausible confounds. [5] The authors noted that no universal standards for laboratory testing protocols exist. [5] A review of US adult-use rules likewise found that states developed their own frameworks, with wide variability in allowable limits, analyte lists and method validation requirements. [6]

Some programs try to improve reliability. California requires its testing labs to hold ISO/IEC 17025 accreditation, and the department describes that standard as requiring labs to show technical competency and reliable results. [1] Accreditation raises confidence in a lab; it does not make every number identical.

Quick comparison

What methods cannot tell you

A measured amount of a compound is a laboratory result. It is not a promise about effects, and it describes one batch tested on one date. For how batch numbers and dates fit in, see Laboratory Variability, Batch Numbers and Test Dates. For the limits of what a report proves, see Why a COA Does Not Prove Clinical Effectiveness.