Scientifically reviewed by Chana Frenkel, Ph.D.
Cannabis testing is the checkpoint standing between a harvest and a dispensary shelf. In state-regulated cannabis markets, products are typically tested by licensed laboratories before sale – screened for contaminants, measured for cannabinoids – with the paperwork to prove it. That much is simple. What the numbers on that paperwork mean is less simple, and it is the subject this magazine is partly named for.
The most useful idea in this guide is also the least intuitive one: a lab result is a measurement, not a property of the plant. Every value on a certificate of analysis was produced by a particular method, run on a particular sample, against a particular calibration. Change any of the three and the number can change too. Holding onto that idea explains almost everything that seems strange about cannabis lab data – from THC values that differ between two competent labs to terpene profiles that shift with the technique used to capture them.
What is cannabis testing for?
Testing does three jobs, and they are worth keeping separate.
Safety. Labs screen for things that should not be in the product: pesticide residues, heavy metals, mold and bacteria, the toxins molds produce, and leftover solvents from extraction. This is the job regulators care about most: state cannabis markets have no unified federal testing and premarket-safety system comparable to those governing federally regulated products, so each state has built its own compliance framework, with laboratories at its center. What labs look for in each contaminant class, and why, is the subject of our cannabis contaminant testing pillar.
Information. Labs also measure what consumers and doctors want to know: cannabinoid potency and, increasingly, the terpene profile. A patient titrating a THC dose needs the milligram numbers to be close to true. A consumer choosing between two chemovars gets more from a terpene panel than from either product’s name.
Marketing. The third job is unofficial but real. THC percentage drives wholesale and retail prices, which gives growers, retailers, and even some labs a reason to prefer generous numbers. That incentive has measurable consequences, and we will get to them.
What do labs actually test cannabis for?
The exact compliance panel depends on the state, but it is assembled from the same categories:
Cannabinoid potency. THC, THCA, CBD, CBDA, and a supporting cast that keeps growing – the NIST mass-spectral library now holds well over a hundred cannabinoids. Potency is the label’s headline number and the market’s pricing signal, which is why it gets its own section below.
Terpene profile. Often optional rather than mandated, but it is the panel that helps describe how a product will smell and taste. How labs separate, identify, and quantify terpenes – and why the method shapes the answer – is the subject of our dedicated pillar on how labs test for terpenes. Volatile compounds also demand their own sample-handling tricks; headspace techniques let a lab sample the vapor above the flower instead of the flower itself.
Pesticides. State screening lists range from a few dozen compounds to several hundred, at action limits that differ by orders of magnitude between jurisdictions.
Heavy metals. Usually lead, cadmium, arsenic, and mercury. Cannabis can take up metals from its growing environment, and permitted concentrations vary substantially between jurisdictions. [4]
Microbials and mycotoxins. Yeast and mold counts, Aspergillus species, and bacterial pathogens on the microbial side; on the toxin side, mycotoxin analysis looks for aflatoxins and ochratoxin A, toxic mold metabolites that can remain even when the mold itself is dead. States disagree philosophically here – some regulate total microbial counts, others only named pathogens.
Residual solvents. For extracts: whatever butane, ethanol, or other process solvents remain in the finished concentrate.
Moisture and water activity. Moisture content affects storage and dry-weight potency calculations, while water activity more directly indicates whether enough available water remains for microorganisms to grow.
One more piece belongs in this picture: remediation. When flower fails a microbial test, many states allow it to be treated and retested rather than destroyed, including by irradiation. Whether that changes the product is a fair question with actual data behind it – we have covered both gamma irradiation and electron-beam treatment in depth.
How do labs measure it?
Most chemical panels begin with the same two-step logic: separate the mixture, then detect what comes out. Heavy metals, microbial contaminants, moisture, and water activity require different approaches.
Separation usually means chromatography. A sample extract is carried through a column that lets different compounds travel at different speeds, so they reach the detector separately. Gas chromatography (GC) vaporizes the sample in a heated inlet; liquid chromatography (LC) analyzes it in solution near room temperature. That difference matters enormously for cannabis: conventional GC conditions can convert THCA into THC during analysis unless the acids are stabilized, while LC can report THCA and THC separately. Properly validated methods should still produce comparable total-THC results within their measurement uncertainty, but they may not print the same THCA-to-THC breakdown. [5]
Detection is a separate choice. Flame ionization detection is the economical workhorse for quantifying known compounds; mass spectrometry adds identification power by fragmenting molecules and matching the pieces against libraries; tandem mass spectrometry buys extra selectivity in complex matrices like cannabis extracts. For heavy metals the instrument changes entirely – inductively coupled plasma mass spectrometry (ICP-MS) measures elemental ions rather than intact molecules. For microbials, laboratories may grow organisms on culture plates or use molecular assays that detect target DNA; because those methods measure different things, they can produce different answers. [7]
Underneath all of it sits an unglamorous dependency: reference materials. Certified reference standards allow laboratories to calibrate their quantitative measurements, while mass-spectral libraries help identify compounds from their characteristic spectra. They are related infrastructure, not the same thing – and expanding both is quiet but important work.
Why do two labs give different THC numbers?
Because potency is measured, not read off the plant. The legitimate reasons stack up fast: trichomes are not distributed evenly, so two samples from the same batch differ before any chemistry happens; results reported on a dry-weight basis move with the moisture measurement; GC and LC handle THCA differently, as above; and total THC must be calculated by converting THCA with a 0.877 multiplier [6] – adding the two numbers directly overstates the THCA-derived portion by roughly 14%, and the overall total by somewhat less when neutral THC is also present.
Then there is the illegitimate layer. Because THC percentage sets price, producers can shop for laboratories known to report generously. Analyses of state seed-to-sale databases have found persistent differences among laboratories even after plausible confounding factors were considered, as well as suspicious clustering of results immediately above the commercially important 20% threshold. [8,9] The full story – methods, studies, and the enforcement actions that followed – is in our deep dive on why two labs give different THC results.
A related honesty problem is older than the potency wars: numbers that circulate for years without anyone checking where they came from. Our audit of cannabinoid boiling points found exactly that – a literature of copied values with surprisingly thin provenance. It is the same lesson at a different scale: ask where the number came from.
What does the law actually require?
There is no single answer, and that is the point. As our lab testing requirements guide lays out, states built their programs independently, without a national cannabis-safety standard – so pesticide lists, microbial philosophies, action limits, sampling rules, and remediation policies all vary by state line. [4] What most mature programs share is the accreditation floor: compliance laboratories are generally required to hold ISO/IEC 17025, the international standard for testing-laboratory competence. [1]
The federal layer is now changing in one specific place: hemp. Most of the new federal definition takes effect December 11, 2026, when hemp moves to a total-THC standard that includes THCA and remains capped at 0.3% by dry weight. Finished hemp-derived cannabinoid products will also be excluded from the definition if they contain more than 0.4 mg per container of total tetrahydrocannabinols, combined with any other cannabinoids that federal health officials determine have similar effects. One provision arrives earlier: beginning November 12, 2026, products containing cannabinoids that the cannabis plant cannot naturally produce are excluded from hemp. [2,3] For laboratories, the shift changes what must be measured to establish whether a product qualifies as hemp; for the market, it closes the loophole that let high-THCA flower ship as paperwork-hemp.
How do you read a cannabis lab result?
A certificate of analysis (COA) is the document all of this produces. A reasonable reader’s checklist:
- Match the batch. The COA’s batch or lot number should match the package in your hand, and the test date should be recent enough to describe it.
- Check the lab. A real, licensed laboratory should be named – not just represented by a logo. Where accreditation is required, verify that it is current and covers the relevant testing scope.
- Read pass/fail by category. A potency-only printout is not a safety panel. Look for the contaminant categories your state requires.
- Understand “ND.” Not detected generally means the analyte was not found above the laboratory’s stated detection or reporting limit – it does not guarantee that none is present. Check the COA’s abbreviations, because labs do not all use ND, LOD, and LOQ identically.
- Do the potency math. Total THC should be THC + (0.877 × THCA). If the document just adds them, be skeptical of the rest of it.
- Treat terpenes as information. Where present, a terpene panel often provides a more useful description of aromatic character than a product name alone – although it remains one of the less standardized parts of the document.
Why it matters
A laboratory certificate is one of the main protections separating regulated cannabis from untested cannabis – and it is exactly as trustworthy as the method, the sample, and the incentives behind it. That is not cynicism; it is how measurement works, in cannabis or anywhere else. The field’s real progress – standard methods, reference materials, accreditation, honest enforcement – is progress on exactly those three fronts. Testing is half of this magazine’s name because it is half of what makes this industry legitimate. The pages linked above go deeper on every piece of it.
References
- ISO/IEC 17025:2017, General requirements for the competence of testing and calibration laboratories. International Organization for Standardization.
- Pub. L. 119-37, §781 (2025) (the federal hemp redefinition); see also Congressional Research Service, IF13136. https://www.congress.gov/crs-product/IF13136
- Continuing Appropriations and Extensions Act, 2027, Pub. L. 119-103, §2019 (2026) (delaying most provisions of the redefinition to December 11, 2026).
- Jameson LE, Conrow KD, Pinkhasova DV, et al. Comparison of State-Level Regulations for Cannabis Contaminants and Implications for Public Health. Environ Health Perspect. 2022;130(9):097001. doi:10.1289/EHP11206
- Pourseyed Lazarjani M, Torres S, Hooker T, Seyfoddin A, Fowlie C, Young O. Methods for quantification of cannabinoids: a narrative review. J Cannabis Res. 2020;2:35. doi:10.1186/s42238-020-00040-2
- U.S. Department of Agriculture, Establishment of a Domestic Hemp Production Program, 7 C.F.R. Part 990 (total THC calculated from THCA using a 0.877 conversion factor).
- McKernan K, Spangler J, Helbert Y, et al. Metagenomic analysis of medicinal Cannabis samples; pathogenic bacteria, toxigenic fungi, and beneficial microbes grow in culture-based yeast and mold tests. F1000Research. 2016;5:2471. doi:10.12688/f1000research.9662.1
- Jikomes N, Zoorob M. The Cannabinoid Content of Legal Cannabis in Washington State Varies Systematically Across Testing Facilities and Popular Consumer Products. Sci Rep. 2018;8:4519. doi:10.1038/s41598-018-22755-2
- Zoorob MJ. The frequency distribution of reported THC concentrations of legal cannabis flower products increases discontinuously around the 20% THC threshold in Nevada and Washington state. J Cannabis Res. 2021;3:6. doi:10.1186/s42238-021-00064-2
