Every legal cannabis product sold in the United States has been tested by a licensed laboratory. The results of that testing are documented in a Certificate of Analysis — a COA — that contains a detailed chemical and safety profile of the product. It is, by a wide margin, the most useful piece of information available to any cannabis consumer.

Almost nobody reads them.

This is partly because COAs are not always easy to access, partly because the format is designed for laboratory professionals rather than consumers, and partly because the industry has trained consumers to focus on a single number — THC percentage — rather than the full picture. This guide breaks down every section of a standard cannabis COA, explains what the numbers mean, and identifies what should concern you.

What Is a Certificate of Analysis?

A COA is an official document produced by a state-licensed testing laboratory that certifies the chemical composition and safety profile of a specific batch of cannabis product. Every COA should include several pieces of identifying information: the name of the testing laboratory, the lab’s license number, the date of testing, the batch or lot number of the product tested, the name of the client (the cultivator, processor, or brand that submitted the sample), and the sample type (flower, concentrate, edible, etc.).

This identifying information is important because it allows you to verify that the COA matches the product you are purchasing. A COA from a different batch, a different product, or a different company is not relevant to the product in your hand — and substituting unrelated COAs is one of the more common forms of lab report fraud.

Most states require that COAs be traceable through the state’s seed-to-sale tracking system, which assigns a unique identifier to every batch of cannabis at every stage of production. If the COA’s batch number matches the batch number on the product label, you can be confident that the lab results apply to that specific product.

Potency Testing

The potency section is what most consumers look at — and it is the most commonly misunderstood part of the COA.

Potency testing reports the concentration of cannabinoids in the product, typically as a percentage of total weight for flower and concentrates, or in milligrams for edibles. The primary cannabinoids tested are Delta-9 THC (or more precisely, Total THC, which accounts for THCA converting to THC when heated), CBD, and often several minor cannabinoids including CBG, CBN, CBC, and THCV.

The Total THC calculation is critical to understand. Raw cannabis flower contains very little active THC — most of it exists as THCA, the acidic precursor that converts to THC through decarboxylation (heating). The standard formula is: Total THC = (THCA x 0.877) + Delta-9 THC. The 0.877 factor accounts for the molecular weight lost during decarboxylation.

This means a flower product listing 25% THCA and 0.5% THC has a Total THC of approximately 22.4%. This is the number that matters for predicting psychoactive potency — not the THCA number alone, though some labels display both.

For edibles, potency is reported in milligrams per serving and per package. State regulations cap serving sizes (typically 5 mg or 10 mg THC per serving) and package sizes (typically 100 mg per package). Check that the COA confirms the product meets these limits and that the actual tested potency matches the labeled potency. Variance of up to 10-15% from the labeled dose is common and generally within regulatory tolerance, but larger discrepancies indicate manufacturing quality control issues.

Terpene Analysis

The terpene section of a COA reports the concentrations of aromatic compounds present in the product. Not all states require terpene testing, but it is available from many labs and increasingly included on premium products.

Terpenes are reported as a percentage of total weight, with individual concentrations typically ranging from 0.1% to 3% for the most abundant terpenes. Total terpene content in flower usually ranges from 1% to 5%.

The terpene profile is arguably more informative than the potency section for predicting the qualitative effects of a cannabis product. Research into the entourage effect suggests that terpenes modulate how cannabinoids interact with the endocannabinoid system — meaning that two products with identical THC percentages but different terpene profiles may produce noticeably different experiences.

The most commonly tested terpenes and their general associations:

Myrcene — the most abundant terpene in most cannabis. Associated with relaxing, sedating effects. Also found in mangoes and hops. Products with myrcene as the dominant terpene tend to be described as “heavy” or “couch-lock” by consumers.

Limonene — citrus-scented terpene associated with elevated mood and stress relief. Also found in citrus peel. Products with high limonene are often described as “uplifting” or “energizing.”

Caryophyllene — spicy, peppery terpene that uniquely binds to the CB2 cannabinoid receptor. Associated with anti-inflammatory effects. Also found in black pepper and cloves.

Linalool — floral terpene associated with calming and sedating effects. The primary terpene in lavender. Products high in linalool are often recommended for anxiety and sleep.

Pinene — pine-scented terpene associated with alertness and memory retention. Also found in pine needles and rosemary. Some research suggests pinene may counteract some of THC’s short-term memory effects.

Terpinolene — floral and herbal terpene found in smaller concentrations. Associated with mildly stimulating effects. Less common as a dominant terpene but significant when present.

Contaminant Testing

The contaminant sections of a COA are the most important for consumer safety — and the sections most consumers skip entirely.

Pesticide screening tests for the presence of chemical pesticides, fungicides, and plant growth regulators. State-mandated panels vary: Colorado tests for 80+ analytes, California tests for 66, and some states test for fewer than 30. A passing result means all tested pesticides were below the state’s action limits — which are the maximum allowable concentrations deemed safe for consumption.

A passing pesticide test is good but not foolproof. Some states’ action limits are more permissive than others, and the tests only screen for the specific compounds on the state’s required panel. A pesticide not on the panel will not be detected even if it is present.

Heavy metals testing screens for lead, arsenic, mercury, and cadmium — contaminants that can be present in soil, fertilizers, or growing media and are absorbed by the cannabis plant during cultivation. All four metals are toxic at sufficient concentrations, and cannabis is a particularly efficient bioaccumulator, meaning it absorbs heavy metals from its environment more readily than many other plants.

A passing heavy metals test is non-negotiable. Failure on any of the four metals should disqualify a product from consideration, regardless of other results.

Microbial testing screens for harmful bacteria (E. coli, Salmonella) and molds (Aspergillus, which can cause serious respiratory infections in immunocompromised individuals). This testing is particularly important for flower products, which are consumed through inhalation and deliver any microbial contaminants directly to the lungs.

Residual solvent testing applies to concentrate products (vape cartridges, dabs, shatter, wax) and screens for chemicals used in the extraction process — butane, propane, ethanol, and others. Residual solvents above safe thresholds can cause respiratory irritation and other health effects. This section is irrelevant for flower products but critical for concentrates.

Mycotoxin testing screens for toxic compounds produced by certain mold species. Even if a microbial test shows no active mold growth, mycotoxins can persist in the product from mold that was present earlier in the cultivation or curing process. Not all states require mycotoxin testing, but its presence on a COA is a positive indicator of testing thoroughness.

How to Spot a Fake or Manipulated Lab Report

Lab report fraud exists. It ranges from unsophisticated (a PDF that has been edited with consumer software) to more subtle (strategic sample selection to ensure a favorable result). Several red flags should trigger additional scrutiny.

No lab license number or accreditation. Every legitimate testing lab carries a state-issued license and often ISO 17025 accreditation. If the COA does not list these, the lab may not be legitimate.

Results that seem too good. Flower testing at 35%+ THC is exceedingly rare in nature, and results consistently above 30% from the same cultivator should be viewed with skepticism. The industry term for inflated potency results is “lab shopping” — the practice of submitting samples to labs known for producing higher numbers.

Missing contaminant sections. A COA that reports potency but omits pesticide, heavy metal, or microbial testing is incomplete. In some states, a potency-only COA may be technically legal for certain product types, but it leaves significant safety questions unanswered.

Mismatched batch numbers. If the batch number on the COA does not match the batch number on the product label, the COA may not apply to the product you are holding.

No QR code or verification link. Many labs provide online verification portals where you can enter a batch number and retrieve the original COA directly from the lab’s system. This is the most reliable way to confirm that a COA has not been altered.

The Lab Testing Controversy

The cannabis testing industry is not without its problems, and consumers should understand the limitations of the system even when it functions as designed.

Lab shopping — sending samples to labs that consistently produce higher potency results — is an acknowledged problem in every major cannabis market. A 2023 investigation by The Cannigma found that the same flower sample submitted to multiple labs in the same state produced THC results ranging from 19% to 27%, a spread that exceeds any reasonable margin of analytical error.

The economic incentive is straightforward: higher THC numbers command higher wholesale and retail prices. Labs that produce higher numbers attract more business. This creates a race to the top that distorts consumer information and undermines the testing system’s credibility.

Several states have implemented or are considering reforms: blind testing (where labs do not know the identity of the client), reference standard audits (where regulators submit known samples to check lab accuracy), and potency variance limits (where results from the same cultivator that swing dramatically between batches trigger investigation).

Until these reforms are widespread, consumers should treat potency numbers as approximate rather than exact, pay more attention to terpene profiles and contaminant results than to THC percentage, and use COAs as one input among many — including personal experience and trusted recommendations — when making purchasing decisions.

The COA is not a perfect document. But it is the best tool available for making informed decisions about what you put in your body, and any consumer who takes cannabis seriously should learn to read one. For a deeper look at the science behind each test method, see our article on how cannabis lab testing works. And if any terminology is unfamiliar, check our cannabis glossary.