How to Read a Cannabis COA (Certificate of Analysis) Like a Lab Analyst
A certificate of analysis (COA) is the lab report behind a cannabis product's numbers — cannabinoid content, usually a terpene profile, and a battery of contaminant screens, all tied to one specific batch. To read one: match the batch number on the document to the one on your package, confirm a named accredited laboratory issued it, scan the contaminant lines for passes, then read the potency knowing 'total THC' is a calculation, not a single measurement. This guide walks through every section of the document, including the red flags that give a fake away. It is education for adults 21 and over — not legal or medical advice.
Key takeaways
If you remember nothing else from this guide, keep these five points. (1) A certificate of analysis is a batch-specific report issued by a testing laboratory, not by the brand, and it only applies to your product if the batch or lot number on the document matches the one printed on your package. (2) 'Total THC' is almost always a calculation — THCA multiplied by 0.877, plus delta-9 THC — so it reads higher than the delta-9 THC row alone, and that is chemistry math, not an error. (3) Contaminant results — pesticides, heavy metals, microbials, mycotoxins, and residual solvents — are judged pass/fail against state limits, and 'ND' means not detected above the lab's reporting threshold, which is not the same thing as a guarantee of absolute zero. (4) The biggest red flags on a COA are a missing batch number, no named laboratory, results that repeat digit-for-digit across different products, and fine print like 'for research purposes only' that quietly disclaims the whole document. (5) In New York's licensed adult-use market, products are tested by laboratories permitted by the Office of Cannabis Management before sale, and it is reasonable to ask a dispensary to show you a product's COA — cannabis.ny.gov is the authority on the current rules. Everything below is general education for adults 21 and over, not legal or medical advice.
What is a cannabis COA, and who issues it?
A certificate of analysis is a formal report issued by a testing laboratory documenting what that lab measured in one specific batch of product: cannabinoid content, often a terpene profile, and a series of contaminant screens, each compared against limits set by state regulators. The lab issues it — never the brand whose name is on the jar. That separation is the whole point: the document exists so that a claim like 'lab-tested' traces back to an independent measurement rather than to marketing copy. BudAbout, which publishes this guide, is run by a former cannabis laboratory analyst — certified in analytical testing and licensed in cannabis extraction (New Jersey and Pennsylvania) — with years inside ISO/IEC 17025-accredited testing labs (ISO/IEC 17025 is the international accreditation standard for testing labs), running cannabinoid potency panels on HPLC (high-performance liquid chromatography) and contaminant screens on GC-MS (gas chromatography-mass spectrometry). That bench-side perspective — generating these documents, not just reading them — sets the standard this guide is held to. Two role clarifications matter before we go further. First, a COA describes a batch, not a brand: the same product line can test differently from harvest to harvest, so the document in front of you speaks only for the batch named on it, at the moment its sample was tested. Second, BudAbout does not lab-test anything — our product checks are visual, on-camera inspections — and every COA we ever discuss was produced by an accredited laboratory for the producer, not by us.
How do you know a COA actually matches your product?
Match three things before you read a single number: the batch or lot number on the document against the one printed on your package, the product name and type, and the dates — a COA is only meaningful if it was issued for the exact batch you are holding, by a named laboratory, reasonably close to when your product was packaged. The batch number is the anchor. On packaging it usually hides in small print near the barcode or on the bottom of the container, labeled 'batch,' 'lot,' or sometimes 'harvest ID'; on the COA it appears in the header block along with the sample ID the lab assigned when the material arrived. If those identifiers do not match, the document is not evidence about your product, no matter how good its numbers look. Watch out for brand-level 'representative' COAs — a single old report a company circulates for every batch of a product. Plant material varies harvest to harvest, so a representative document is a weak substitute for a batch-matched one. Dates matter too: a test performed long before your package date describes flower that has since kept drying and changing, and volatile compounds like terpenes fade over months. A COA from a named lab, matched to your batch, dated in the same season as your package — that is the document worth reading.
Edible COAs flip to milligrams because serving-level dosing is what their labels regulate.
How do you read the cannabinoid potency section?
The potency section is a table of cannabinoids — THCA, delta-9 THC, CBD, CBDA, CBG, CBN and others — each with a result, usually expressed as a percentage by weight for flower and as milligrams per serving and per package for edibles. Find the rows, note the units, and check which figures are measured versus calculated. On the instrument side, a modern potency lab measures the acid forms (THCA, CBDA) and the neutral forms (delta-9 THC, CBD) as separate compounds, because chemically they are separate compounds — which is why an honest COA lists them on separate lines rather than as one blended 'THC' number. Units trip people up less once you know the conversion: one percent by weight equals ten milligrams per gram, so flower reporting 20% total THC is carrying about 200 mg per gram of material. Edible COAs flip to milligrams because serving-level dosing is what their labels regulate. Expect most rows in the panel to read 'ND' — a lab tests for a dozen or more cannabinoids and most samples only carry meaningful amounts of a few. That is normal, not suspicious. The rows that deserve your attention are the ones with numbers, the 'total' lines calculated from them, and the units they are printed in.
Why is 'total THC' higher than the delta-9 THC row?
Because total THC is a calculation: THCA multiplied by 0.877, plus delta-9 THC. Raw flower carries most of its THC as THCA, and when THCA is heated it converts to delta-9 THC through a reaction called decarboxylation, shedding part of its mass as carbon dioxide along the way. The 0.877 factor is the molecular-weight adjustment for that lost mass. Walk through an example: flower measuring 22% THCA and 0.8% delta-9 THC has a total THC of 22 x 0.877 + 0.8, or roughly 20.1%. Notice the total is lower than the raw sum of the two rows — the 0.877 factor is doing real chemistry work, accounting for the fact that a gram of THCA cannot become a full gram of delta-9 THC. The same arithmetic applies on the CBD side: total CBD equals CBDA x 0.877 plus CBD. This is why the front of a label and the potency table can both be 'right' while showing different numbers — the label usually quotes the calculated total, while the COA shows you the measured ingredients behind it. A document that shows the acid and neutral forms separately, with the totals derived from them, is showing its work. One that prints only a single suspiciously round 'THC%' with no underlying rows is asking you to take arithmetic on faith.
Why might the label or menu number differ from the COA?
Rounding, batch-to-batch variation, moisture changes, and time. The number printed on a label or a dispensary menu is producer-reported — pulled from a COA at packaging time — while the flower in the jar keeps drying and its chemistry keeps drifting afterward. Small differences are normal; large ones deserve a question. Start with sampling: plant material is not homogeneous, and the buds a lab tested are not the exact buds in your jar, so some spread between the reported figure and any given gram is built into the exercise. Moisture matters too — as flower dries, the same amount of THC becomes a slightly larger percentage of a lighter sample. Then there is the less innocent explanation: regulators and researchers in several states have documented cases of inflated potency reporting, where numbers were nudged upward because high-THC listings sell. We are not accusing any particular producer of anything — most differences have boring explanations — but the pattern is documented widely enough that a healthy shopper treats a THC percentage as an approximate, producer-reported figure rather than a precision instrument. On BudAbout, every THC number you see is labeled producer-reported for exactly this reason: we did not measure it, the producer's lab did, and the honest way to present it is as their reported figure, not a verified fact about your specific gram.
Now the honest boundary: you will find charts all over the internet mapping individual terpenes to feelings and outcomes.
What does the terpene panel tell you?
The terpene panel lists the aromatic compounds the lab measured — beta-myrcene, limonene, beta-caryophyllene, pinene, linalool and others — each as a percentage by weight. Read it as a smell-and-flavor preview: it tells you whether a product leans citrusy, piney, peppery, floral, or earthy-sweet, and nothing more than that. Terpene testing is often optional rather than required, so plenty of legitimate COAs simply omit the panel; its absence is not a red flag the way a missing batch number is. When the panel is present, the practical move is to read the top two or three entries, because they dominate what your nose will find: a jar led by limonene will read bright and citrus-peeled, one led by beta-myrcene rounder and earthy-sweet, one heavy in beta-caryophyllene warmer and peppery. Total terpene content for flower commonly lands in the low single digits of a percent, and higher totals generally mean a louder jar. Now the honest boundary: you will find charts all over the internet mapping individual terpenes to feelings and outcomes. Research has explored those questions, but controlled human evidence at flower-level amounts is limited and unsettled, and a COA cannot tell you how a product will make you feel. We treat terpene rows strictly as flavor notes — a vocabulary for shopping by smell, not a prediction of anything.
What are the contaminant panels, and what does 'pass' mean?
Contaminant panels screen a batch against state limits for pesticides, heavy metals (lead, arsenic, cadmium, mercury), microbials (organisms like salmonella, E. coli, and certain molds), mycotoxins (toxic byproducts some molds produce), and — mainly for concentrates — residual solvents left over from extraction. 'Pass' means every result came in under the state's action limit for that compound, not that the result was zero. This is the part of the document most worth a shopper's attention, and also the part most people skip because it is a wall of unfamiliar names. You do not need to know the compounds; you need to read the verdict column. Each panel typically shows the compound, the state's limit, the measured result (often 'ND'), and a pass/fail status, with concentrations in parts per million or parts per billion — units small enough that 'detected' and 'concerning' are very different statements, which is exactly why the limit column exists. Residual solvent panels apply to extracted products like vapes and concentrates because solvents such as butane or ethanol are used in manufacturing and must be removed to below limits; flower that was never extracted has nothing to purge, so a flower COA may legitimately omit that panel. You may also see moisture content and water activity rows on flower — those are quality and mold-risk indicators the state also caps. The whole contaminant story compresses to one question: does every panel present say pass, on a document that actually matches your batch?
What do ND, LOQ, and LOD actually mean?
ND means 'not detected' — the compound was not found above the lab's reporting threshold. LOQ is the limit of quantitation, the smallest amount the method can reliably put a number on; LOD, the limit of detection, is the smallest amount it can reliably see at all. None of these mean 'absolute zero.' Every analytical method has a floor below which it cannot distinguish a trace of a compound from background noise, and honest labs state those floors on the document instead of pretending to measure the unmeasurable. That is why ND is always ND relative to a threshold. Occasionally you will see a result reported as '<LOQ' — the instrument saw a trace above the detection limit but below the level it can reliably quantify. On a contaminant panel, a trace below the state's action limit is still a pass; the limits exist precisely because 'present in any amount whatsoever' is not a standard any agricultural product on earth could meet. From the analyst's side of the bench, the practical standard is this: ND at a sensible, stated LOQ is what 'clean' looks like on paper. A COA that lists its LOQs is being transparent about what its numbers mean; one that reports a wall of perfect zeros with no thresholds stated is glossing over how measurement actually works — real instruments report thresholds, not certainties.
Any one of these flags is a reason to pause; two or more and the document is telling you more about the seller than about the product.
What are the red flags of a fake or worthless COA?
The five biggest: no batch or lot number, no named and accredited laboratory, results that repeat digit-for-digit across different products, potency that looks too good to be true sitting beside a spotless pass on everything, and fine print like 'for research purposes only' that quietly disclaims the whole document. Take them in order. A COA with no batch number cannot be tied to any product, which makes it decorative. A document with no lab name, no accreditation mark, no address or contact information cannot be verified — a real report says who did the work, and many labs will confirm a report's authenticity if you contact them or use their online verification portal, which is also the check that catches edited PDFs. Copy-paste results are a quieter tell: plant material varies, so if a brand's 'different' strains all show identical cannabinoid and terpene numbers to the second decimal, someone duplicated a document rather than testing a batch. Implausible potency deserves skepticism too — flower reporting in the high thirties or forties of percent total THC is far outside what verified samples typically show, and extraordinary numbers paired with a perfect contaminant record on a document nobody can verify is the classic profile of a COA built to sell rather than to inform. Finally, read the disclaimers: 'R&D only,' 'not for compliance purposes,' or 'informational use' means the lab itself is telling you this was not a regulatory compliance test. Any one of these flags is a reason to pause; two or more and the document is telling you more about the seller than about the product.
How does COA testing work in New York?
In New York's licensed adult-use market, cannabis products must be tested by laboratories permitted by the Office of Cannabis Management (OCM) before they reach a dispensary shelf, against the state's rules for contaminants, potency reporting, and labeling. The specifics — which panels are required, at what limits — are set by the OCM and evolve over time, so treat cannabis.ny.gov as the authority on current requirements rather than any blog, including this one. The structural point for a shopper is the difference between the licensed and unlicensed markets. A product on a licensed dispensary's shelf sits at the end of a chain that includes mandatory testing by a state-permitted lab; a product in an unlicensed shop carries no such assurance, whatever its packaging claims, and unlicensed sellers have every incentive to dress products in official-looking paperwork. That is general information about how the system is built, not legal advice about any particular purchase. Licensed-market labels in New York also carry markings that help you connect product to paperwork — license details and, commonly, a QR code or batch identifier that leads toward the underlying test information. If you want to go deeper on the label side, our separate guide to reading a New York cannabis label covers those markings; this guide's job is the document behind them.
Can you ask a dispensary to see the COA?
Yes — asking is normal and reasonable. Many brands print a QR code on the package that links to the batch's COA, some post their COAs on their websites searchable by batch number, and dispensary staff can often pull one up at the counter if you ask. How readily the document appears varies by shop and by brand, and that responsiveness is itself information. You do not need to perform a full audit at the register. A thirty-second counter check covers the essentials: does the batch number on the document match the package, is there a named laboratory on it, and do the contaminant panels say pass. That is a completely reasonable ask in a regulated market, and no good budtender is offended by it — the licensed system's entire sales pitch over the unlicensed market is that this paperwork exists. If nobody can produce a COA for a product after a genuine attempt, that is not proof of anything sinister, but it is a data point, and you are always free to spend your money on the batch whose paperwork showed up. For the current rules on what New York requires of dispensaries and producers, check cannabis.ny.gov directly — requirements change, and the state's own site outranks anyone's summary of it, ours included.
That covers most of what the document can honestly tell you.
How should you actually use a COA when shopping?
Use it as a sixty-second routine: match the batch number to your package, confirm a named accredited lab issued it, scan the contaminant panels for passes, note the total THC with the 0.877 calculation in mind, and read the top terpene rows as a flavor preview. That covers most of what the document can honestly tell you. Just as important is what a COA cannot tell you. It says nothing about freshness, cure, trim, or handling — a batch can pass every panel and still arrive as dry, flavorless flower, because the document describes chemistry at test time, not condition at sale time. It cannot tell you how a product will make you feel, and it is not a substitute for your own judgment about what you enjoy. The strongest position for a shopper is holding both kinds of evidence at once: the lab's paperwork for what is measurable, and your eyes and nose for what is not. That division of labor is also exactly where BudAbout sits — our checks are visual, on-camera inspections of condition and quality, we do not lab-test anything, and every potency figure we display is producer-reported from documents like the ones this guide teaches you to read. This article is general education for adults 21 and over, not legal or medical advice; for New York's current testing and labeling rules, cannabis.ny.gov is the source that counts.
FAQ
Do NYC dispensaries have to show you a COA?
Products in New York's licensed adult-use market are tested by OCM-permitted laboratories before sale, but how the document itself reaches shoppers varies — some brands link it from a QR code on the package, some post COAs online by batch number, and staff can often pull one up on request. Ask the shop, and check cannabis.ny.gov for the current testing and labeling rules. This is general information, not legal advice.
What does ND mean on a COA?
ND means 'not detected' — the lab did not find the compound above its stated reporting threshold. It is not a claim of absolute zero, because every analytical method has a floor below which it cannot reliably measure. ND at a sensible, stated limit of quantitation is what a clean result looks like on a real lab report.
Why is total THC higher than the delta-9 THC line?
Because total THC is calculated as THCA x 0.877 plus delta-9 THC. Raw flower holds most of its THC as THCA, which converts to delta-9 THC when heated, losing some mass as carbon dioxide — the 0.877 factor accounts for that lost mass. So the total reads higher than the delta-9 row alone, and slightly lower than simply adding the two rows.
How do I check that a COA matches my product?
Match the batch or lot number on the document to the one printed on your package — usually in small print near the barcode — then confirm the product name, a named laboratory, and a test date reasonably close to your package date. A COA without a matching batch number is not evidence about the product in your hand.
Can a COA tell me how a product will make me feel?
No. A COA is a chemistry snapshot of a batch sample — cannabinoid and terpene content plus contaminant screens — not a prediction of anyone's experience. We read terpene rows as aroma and flavor information only, and nothing in this guide is medical advice. It is educational content for adults 21 and over.
Does BudAbout lab-test products or issue COAs?
No. BudAbout performs visual, on-camera product checks and never lab-tests anything. Every COA we discuss was issued by an accredited laboratory for the producer, and every potency figure on our site is producer-reported. This guide exists to help you read those documents yourself, not to substitute for them.