Cannabis testing in the United States has no single national standard. Requirements are written state by state, methods differ between laboratories inside the same state, and reports vary enough in format that comparing two of them takes effort.
That fragmentation is the root of most complaints about testing reliability. Understanding where cannabis testing standards come from — and where they do not exist — explains a lot about why results behave the way they do.
Who Sets the Rules
State regulators are the primary authority. Each cannabis programme defines which analyses are required for which product types, what the action limits are, and what a laboratory must do to be licensed. Because these decisions are made independently, they diverge — a pesticide permitted in one state is prohibited in the next, and action limits for the same compound can differ by an order of magnitude.
Accreditation bodies assess technical competence against ISO/IEC 17025. Accreditation evaluates whether a laboratory can produce valid results using the methods within its scope. It does not prescribe which methods a state should require.
Standards organisations — AOAC International, ASTM, the US Pharmacopeia among them — develop and validate analytical methods that regulators and laboratories can adopt. This is the slow, unglamorous work that eventually reduces variation. Adoption is voluntary and uneven.
Federal agencies are largely absent. Cannabis’s federal status keeps the FDA and USDA out of the space in the way they operate for food and pharmaceuticals, which is precisely why the vacuum exists.
What Standardization Would Actually Fix
Inter-laboratory variation. When methods are prescribed rather than left to each laboratory, the range of defensible choices narrows and results converge. Much of the disagreement described in the limits of current testing methods is method choice rather than error.
Lab shopping. Shopping only pays if results differ enough between laboratories to be worth chasing. Standardised methods reduce the spread and therefore the incentive.
Comparability across states. Multi-state operators currently maintain different testing regimes per market for the same product. Harmonisation would remove duplicated work and make data portable.
Consumer interpretation. A uniform report format would let anyone compare two products without decoding two layouts.
What a Transparent Report Contains
Formats differ, but the information that makes a certificate usable does not. A report worth relying on states:
- Sample identification — batch or lot number matching the product, sample date, receipt date, analysis date
- Methods — which analytical method was used for each analysis, referenced to a published standard where one exists
- Results with units — clearly labelled, with the basis stated (dry weight or as-received) for potency
- Action limits — the threshold applied to each analyte, and the jurisdiction it comes from
- Limits of detection and quantitation — without these, “not detected” cannot be interpreted
- Pass/fail status per analysis — not only an overall verdict
- Laboratory identification — name, licence number, accreditation status and scope
- Authorised signature — a named person taking responsibility for release
A certificate missing detection limits, methods or an accreditation reference is not necessarily wrong. It is harder to evaluate, and that is a meaningful difference. Our walkthrough of an actual report goes section by section.
What Happens When Standards Are Not Met
Consequences for laboratories vary by jurisdiction and by how the failure surfaced.
Routine audit findings usually trigger corrective action plans with a deadline. Failed proficiency tests typically require investigation, root cause analysis and often re-testing under observation. Serious or repeated findings lead to suspension of specific test categories or of the licence entirely. Where results were falsified, invalidation of previously issued certificates and product recalls follow — and in some jurisdictions, criminal referral.
The variable that matters most is whether enforcement is public. Where regulators publish suspensions and findings, buyers can factor them into laboratory selection and the market corrects. Where enforcement happens privately, a laboratory can be under corrective action while continuing to win work from clients who have no way of knowing.
Proficiency Testing as the Check
Proficiency testing is the mechanism that makes standards enforceable in practice. Laboratories receive samples with known values, analyse them blind, and report results that are compared against the reference.
It is the only way to distinguish accuracy from internal consistency. A laboratory can be highly repeatable and consistently wrong, and only an external reference reveals that.
Voluntary programmes exist and are useful. Mandatory blind programmes — where the laboratory does not know a sample is being assessed — are considerably stronger, because they measure ordinary performance rather than performance under observation. States that have introduced blind proficiency testing have found problems that voluntary programmes did not surface.
Where This Is Heading
Progress is real but slow. AOAC and ASTM continue to publish validated methods, and adoption widens gradually. Several states have converged on similar requirements, mostly by copying each other. Blind proficiency testing is expanding. Some multi-state operators are standardising internally ahead of regulation because it is cheaper than maintaining parallel systems.
Federal reform would accelerate all of it, because federal oversight would bring a single set of methods and limits. Until then, harmonisation depends on states and standards bodies converging voluntarily. More on the trajectory in the future of cannabis lab testing and on the current patchwork in how testing rules differ across states.
What Operators Can Do Now
Waiting for standardisation is not a strategy. In the meantime:
- Work with an accredited laboratory and confirm the accreditation scope covers your analyses and matrices.
- Ask which published methods are used, and whether they are validated for your product types.
- Ask about proficiency testing participation and results.
- Keep your own records so you can spot drift in your data over time.
- Stay with one laboratory where you can — comparability within your own data set is worth more than chasing individual results.
The selection criteria in full are in how to choose a cannabis testing laboratory.