Edited by Nani Frenkel, chief editor
A cannabis lab doesn’t run one test. It runs a series of tests, each using a method suited to a different question. This guide to cannabis testing methods explains what each technique actually measures, in plain terms, and what a passing result can and can’t tell you. For the big picture of why products are tested at all, start with our complete guide to cannabis testing.
Why Does One Batch Need So Many Tests?
Because no single instrument can answer every question a regulator asks. California, for example, requires a menu of tests that varies by product type: cannabinoids, foreign material, heavy metals, microbes, mycotoxins and pesticides, plus moisture, water activity and residual solvents where applicable. Terpenes may be analyzed if requested.[1]
Those questions need very different tools. Measuring THC is a matter of percent. Finding a regulated pesticide means spotting trace amounts. Checking for mold is biology, not chemistry. A certificate of analysis is the sum of all those separate measurements.
How Do Labs Separate What’s in the Sample?
The question: what’s in here, and how much of each? The answer: chromatography, which pulls a mixture apart so each compound can be measured on its own.
Think of it as a race. The sample is dissolved and pushed through a long, narrow tube called a column. Different molecules move through it at different speeds, so they reach the finish line, the detector, at different times. Each compound gets its own moment to be counted.
There are two main kinds. Gas chromatography (GC) turns the sample into a vapor and runs it through a heated column. Liquid chromatography (LC), usually as high-performance liquid chromatography, or HPLC, keeps everything in liquid at milder temperatures.
That difference matters most for potency. In the raw plant, most THC exists as THCA, an acid form. The heat of a gas chromatograph’s injection port converts THCA into THC, unless the lab first chemically protects it, a step called derivatization.[2] So GC can blur the line between the two. Liquid chromatography avoids that heat problem, which is why it is the usual choice for cannabinoids.[2] California now requires labs to use a standardized HPLC method for potency in dried flower and non-infused pre-rolls.[3] Our explainers cover how HPLC measures cannabis potency and how potency testing works.
The column is where the separation happens, so its material, length and coating decide which compounds come apart cleanly and which overlap. Choosing the right column is a large part of a method, and one reason two labs’ methods may differ, except where regulators mandate a standardized method.
How Do Labs Identify What They’ve Separated?
The question: exactly which compound is this? The answer: mass spectrometry (MS), which identifies molecules by the mass of their charged forms.
If chromatography is the race, mass spectrometry checks the runners’ molecular identities as they cross the line. An ion source turns molecules into charged particles, and the instrument separates those ions by their mass-to-charge ratios. Many methods also fragment selected ions; the resulting pattern helps confirm identity. In tandem mass spectrometry (written MS/MS), a selected ion is fragmented and characteristic product ions are measured, providing a stricter identity check.
For pesticides and mycotoxins, that certainty is essential. California assigns each regulated pesticide a numerical action level. Through March 2028, laboratories must establish a limit of quantitation at or below that level, and a sample passes when no measured residue exceeds the applicable limit.[4] AOAC International has published performance requirements for measuring selected pesticide residues in dried cannabis, and accepts any technique that meets them.[5] One commercial workflow designed around those requirements and North American regulatory panels covered 105 pesticides plus mycotoxins, using both LC-MS/MS and GC-MS/MS because the compounds differ in which technique measures them best.[6] Mycotoxins, the poisons some molds leave behind, are also tested at trace levels: California checks four aflatoxins and ochratoxin A.[7]
For potency, mass spectrometry is optional. THC and CBD are present in large amounts, and California’s standard flower method runs on HPLC with light-based detection.[3] Our short primer on mass spectrometry in cannabis labs explains the basics, and two deeper pieces show it at work on 93 terpenes at once and on more than 1,000 pesticides.
How Do Labs Measure Terpenes and Leftover Solvents?
The question: which volatile aroma compounds are present, and is any extraction solvent left behind? The answer: gas chromatography.
Terpenes and solvents such as butane, propane and ethanol evaporate easily, which is exactly what GC needs. For solvents, many labs use a technique called headspace: the sample is warmed in a sealed vial, and only the vapor above it goes into the instrument. That keeps the heavy, sticky parts of an extract out of the column. Our piece on headspace GC for terpenes and solvents walks through it, and gas chromatography for cannabis analysis covers the instrument.
California requires residual solvent testing for cannabis products and pre-rolls.[8] Terpene testing is not routinely required: terpenoid analysis is performed if requested.[9]
How Do Labs Find Heavy Metals?
The question: is there lead in this? The answer: inductively coupled plasma mass spectrometry, ICP-MS for short.
The acid-digested sample is sprayed into a very hot plasma, which atomizes and ionizes its elements. The mass spectrometer separates the resulting ions by mass-to-charge ratio, and their signals are compared with calibration standards.[10] California tests for four: arsenic, cadmium, lead and mercury, with separate limits for products that are inhaled and products that are not.[11]
Metals can come from two places: the plant material itself, and, in vape products, the hardware. A 2025 review of the research concluded that parts of cannabis vape devices can release metals such as nickel, chromium and lead into the oil and aerosol, depending on the device’s condition, age, temperature and the oil, on top of metals that came from the cannabis.[12] Our guide to cannabis contaminant testing covers all the contaminant classes in more depth.
Is There a Faster Way?
The question: can we get an answer in seconds? The answer: spectroscopy, which shines light on a sample and reads what it absorbs.
Techniques such as near-infrared (NIR), Fourier-transform infrared (FTIR) and Raman spectroscopy need little or no sample preparation and can give a result almost instantly. These methods do not ordinarily separate cannabinoids before estimating them. Quantitative NIR and FTIR results usually depend on chemometric models calibrated against samples previously measured by a reference method such as chromatography, so they are only as good as that calibration.
Our review of a handheld NIR study on hemp flour shows the catch: the method relied on gas chromatography data to build its model, and it needed far more validation before it could be trusted routinely.[13] Spectroscopy is useful for screening and process checks. In California, it cannot replace the standardized HPLC method required for potency testing of dried flower and non-infused pre-rolls.[3]
Are Target Microbes Present, and Can They Grow?
The question: are there dangerous microbes? The answers: plating or qPCR, plus a separate test for mycotoxins.
California requires testing for Salmonella and Shiga toxin-producing E. coli in all products, and for four disease-causing Aspergillus molds in products that are inhaled.[14] Labs use two approaches. Plating spreads a sample on a growth medium and counts organisms capable of growing under the test conditions. qPCR (quantitative polymerase chain reaction) detects selected DNA sequences rapidly and specifically, but ordinary qPCR does not necessarily distinguish viable organisms from residual DNA left by dead cells. Our article on qPCR vs. plating compares them.
Killing mold does not necessarily remove the toxins it previously produced. That is why mycotoxins get their own test, typically by mass spectrometry, separate from the microbial one.[6,7]
Which Tests Cause the Most Trouble?
The plainest ones. Before any preparation begins, the laboratory must collect a representative sample from the batch. Homogenization can distribute the material that was collected evenly, but it cannot correct a biased or unrepresentative sample.[1]
Foreign material is largely a visual examination. Before homogenizing the sample, the laboratory examines dried flower inside and out for excessive dirt, mold, insect fragments, hair, mammalian excreta and embedded material.[15] This inspection must happen first because grinding would destroy the evidence it is meant to find.
Except for foreign-material inspection, which happens first, the collected sample increments must be homogenized before analysis.[1] Poor mixing, grinding or weighing at this stage feeds straight into the final number, however good the instrument.
Moisture and water activity sound similar but measure different things. Water activity can fail a batch; moisture content is reported for dried flower. Moisture content measures how much water the sample contains. Water activity measures how available that water is to support microbial growth. California requires the moisture percentage to be reported for dried flower, but water activity—not moisture percentage—determines whether the sample passes this part of testing: the limits are 0.65 for dried flower and pre-rolls and 0.85 for solid edibles.[16]
Calibration matters as much as preparation. In a 2021 NIST exercise, labs measuring the same hemp oils mostly agreed with one another, yet more than half fell outside NIST’s narrow tolerance range, a pattern NIST said often points to calibration bias.[17] That is why two labs can give different THC numbers for the same product.
Thermal analysis tracks how a material changes as it is heated: differential scanning calorimetry (DSC) measures heat flow and transitions such as melting, and thermogravimetric analysis (TGA) measures mass lost during heating. It does not appear on California’s list of required tests,[1] so it belongs to research and product development rather than compliance.
What Does a Passing Test Prove?
Less than many people think. A test result is a measurement made by one method, on one sample, at one time. It is only as good as the method, the sample and the lab.
Quantitative chemical methods commonly have two important reporting limits. The limit of detection (LOD) is the smallest amount it can distinguish from background. The limit of quantitation (LOQ) is the lowest amount it can measure with acceptable performance. So “not detected” means “below what this method can see,” not “none at all.” A more sensitive method might find something.
Standards bodies such as AOAC International and ASTM set performance requirements that methods should meet; our overview of current standard methods for cannabis testing explains them. Lab accreditation adds another layer, which we cover in our guide to ISO 17025 accreditation.
Cannabis Testing Methods at a Glance
| Technique | Question it answers | California compliance use | Our deeper article |
|---|---|---|---|
| Liquid chromatography (HPLC) | How much THC, CBD and other cannabinoids? | Required analyte class; HPLC method mandated for dried flower and non-infused pre-rolls | How HPLC measures potency |
| Gas chromatography (GC) | Which terpenes and leftover solvents? | Commonly used for required solvent testing; instrument not specifically mandated. Terpenoid analysis if requested | GC for cannabis analysis |
| Mass spectrometry (LC-MS/MS, GC-MS/MS) | Are there pesticides or mycotoxins, even in trace amounts? | Commonly used for required pesticide and mycotoxin testing; instrument not specifically mandated | Mass spectrometry primer |
| ICP-MS | Is there arsenic, cadmium, lead or mercury? | Commonly used for required heavy-metals testing; instrument not specifically mandated | Contaminant testing guide |
| Plating and qPCR | Are culturable organisms or target microbial DNA present? | Commonly used for required microbial testing; method not specifically mandated | qPCR vs. plating |
| Visual inspection | Is there dirt, mold, insect fragments or hair? | Required foreign-material test, done before homogenizing | — |
| Moisture and water activity | How much water is present, and how available is it to microbes? | Water activity pass/fail for flower, pre-rolls and solid edibles; moisture reported for flower | — |
| Spectroscopy (NIR, FTIR, Raman) | A quick estimate, in seconds | Screening only; cannot replace the mandated HPLC potency method for dried flower and non-infused pre-rolls | NIR on hemp flour |
| Thermal analysis (DSC, TGA) | How does the material behave when heated? | Not a required test | — |
Requirements shown are California’s; other states differ.
References
- Cal. Code Regs. tit. 4, § 15714, Required testing. Via Legal Information Institute, Cornell Law School.
- García-Valverde MT, Sánchez-Carnerero Callado C, Díaz-Liñán MC, Sánchez de Medina V, Hidalgo-García J, Nadal X, Hanuš L, Ferreiro-Vera C. Effect of temperature in the degradation of cannabinoids: from a brief residence in the gas chromatography inlet port to a longer period in thermal treatments. Frontiers in Chemistry. 2022;10:1038729.
- California Department of Cannabis Control. Standard cannabinoid test method and operating procedures: FAQs.
- California Department of Cannabis Control. DCC-2025-03-R: Pesticide testing (amended Cal. Code Regs. tit. 4, § 15719; effective October 1, 2026); modified regulation text.
- AOAC International. Standard method performance requirements (SMPRs) for identification and quantitation of selected pesticide residues in dried cannabis materials. AOAC SMPR 2018.011. 2018.
- Waters Corporation. Determination of pesticide and mycotoxin residues in dried cannabis flower: LC-MS/MS and GC-MS/MS methodology to meet the recommended AOAC regulatory requirements for US states and Canada. Application note 720006780; 2020.
- Cal. Code Regs. tit. 4, § 15721, Mycotoxin testing. Via Legal Information Institute, Cornell Law School.
- Cal. Code Regs. tit. 4, § 15718, Residual solvents and processing chemicals testing. Via Legal Information Institute, Cornell Law School.
- Cal. Code Regs. tit. 4, § 15725, Terpenoid testing. Via Legal Information Institute, Cornell Law School.
- Agilent Technologies. AOAC approved method for heavy metals testing in cannabis.
- Cal. Code Regs. tit. 4, § 15723, Heavy metals testing. Via Legal Information Institute, Cornell Law School.
- Gaur S, Agnihotri R. Heavy metals in cannabis vapes and their health implications: a scoping review. Scientific World Journal. 2025;2025:9529544.
- Risoluti R, Gullifa G, Battistini A, Materazzi S. Monitoring of cannabinoids in hemp flours by MicroNIR/Chemometrics. Talanta. 2020;211:120672. Reviewed in Lupoi JS, Spectroscopic analysis of hemp flour, Terpenes and Testing Magazine.
- Cal. Code Regs. tit. 4, § 15720, Microbial impurities testing. Via Legal Information Institute, Cornell Law School.
- Cal. Code Regs. tit. 4, § 15722, Foreign material testing. Via Legal Information Institute, Cornell Law School.
- Cal. Code Regs. tit. 4, § 15717, Moisture content and water activity testing. Via Legal Information Institute, Cornell Law School.
- Abdul-Rahman M, Phillips MM, Wilson WB. Cannabis Quality Assurance Program: Exercise 1 final report. NIST IR 8385. July 2021.
