Cannabis Testing Standards: Why They Vary Wildly by State

October 9, 2026

Photo via Pexels.

Pull a lab report from a dispensary in Denver and one from a dispensary in Albany for the same strain name, and you might be looking at two different definitions of what "tested" even means. That's not a glitch. It's the design. Cannabis is still federally restricted, which means there's no binding national baseline sitting underneath state programs the way there is for, say, bottled water or infant formula. Every state has had to build its own testing rulebook from the ground up, usually under legislative deadlines, often with input from whatever labs and lobbyists showed up to the hearings.

The result is a patchwork with real teeth. Depending on the state, a required analyte panel might run as lean as 15 compounds or balloon past 90, covering everything from residual solvents to a long list of named pesticides to heavy metals and mycotoxins. That spread isn't cosmetic -- it determines what gets caught before it reaches a shelf and what slides through with a clean-looking label. And when the system breaks, it breaks publicly: fraud scandals in New York and Michigan over the past year have shown exactly what happens when a lab decides compliance costs are optional. For growers shipping genetics or flower across state lines, and for buyers trying to make sense of a THC number, understanding this variation isn't academic. It's the reason the same cultivar can read 24% THC in one state and 31% in another, with both numbers carrying an official-looking stamp.

The Patchwork: No Federal Floor, 40-Plus Different Rulebooks

Photo by Léster Lau via Pexels.

Start with the basic legal fact that shapes everything else: cannabis sits in a federal gray zone, so no national agency has ever issued a binding national testing standard the way exists for food safety. That absence didn't stop legalization -- it just meant each state legislature and regulatory body had to write its own panel of required tests, usually borrowing pieces from neighboring states or from industries that seemed vaguely analogous. The outcome is 40-plus separate rulebooks, built at different times, by different staff, with different political pressures behind them.

The gap in rigor between states is not subtle. Some states require testing for as few as 15 analytes -- enough to check potency and maybe a short pesticide list. Others require more than 90, covering a long roster of pesticides, heavy metals, residual solvents, microbial contaminants, and mycotoxins. California's Department of Cannabis Control takes an additional structural step: retailers are only permitted to stock products tested by labs licensed through the state's Department of Food and Agriculture, tying retail compliance directly to lab licensure status rather than just a passing result.

Colorado's Marijuana Enforcement Division goes at the problem from the lab side, requiring every testing facility in the state to hold ISO 17025 accreditation -- the same international standard used for calibration and testing labs across industries like environmental monitoring and metrology. That's a meaningfully higher bar than simply registering as a cannabis lab with a state agency. Massachusetts' Cannabis Control Commission layers its requirements by category, mandating potency, pesticide, microbial, and mycotoxin testing before any product reaches a retail shelf, with specific action limits for each.

What this variability produces, in practice, is a two-track industry. Some labs chase multi-state accreditation and build genuinely rigorous operations because their clients sell in strict states. Others operate in the loosest jurisdictions where a lighter panel is legal and cheaper, and compete largely on turnaround time and price. Both kinds of labs can print a lab report that looks identical to a buyer who doesn't know what to check. Tested safe means something very different depending on which rulebook was in play.

What Changed in 2025 and 2026

Photo by 2H Media via Unsplash.

The regulatory ground hasn't been static. Minnesota activated its full testing requirement in October 2025, bringing microbial contaminants, mycotoxins, heavy metals, and pesticides under mandatory screening for the first time in a program that had previously leaned on a narrower panel. That's a significant jump for a state that only recently stood up adult-use sales, and it signals that newer markets are entering with stricter expectations than some of the states that legalized a decade earlier.

New York moved in a more targeted way, updating its pathogen standards in February 2025 to tighten absence requirements for Aspergillus, Salmonella, and STEC (Shiga toxin-producing E. coli). Aspergillus in particular has been a growing focus nationally because of documented cases of invasive aspergillosis in immunocompromised patients linked to contaminated cannabis, so a state revisiting its absence standard for that specific mold is addressing a known clinical risk, not just checking a regulatory box.

Not every state is moving at the same pace, though. Maine's proposed LD 1847 would set testing requirements that, if passed as currently written, would rank among the most permissive in New England -- a notable gap given that Massachusetts, right next door, runs one of the more comprehensive panels in the country. That contrast is a useful illustration of how little geography predicts regulatory rigor; two adjacent New England states can land on opposite ends of the strictness spectrum.

The biggest structural shift on the horizon isn't coming from any state at all. In 2026, the DEA's Diversion Control Division begins regulating testing for Schedule III medical cannabis products, which introduces a federal layer -- Medical Marijuana Analytical Labs -- sitting alongside, not replacing, existing state frameworks. That creates an entirely new category of federally regulated lab work operating in parallel with state-licensed adult-use testing, and it's reasonable to expect friction as the two systems bump into each other over the next few years.

Put these together and the direction is unmistakable: toward stricter, more harmonized pathogen and contaminant rules, not looser ones. States that legalize later are generally writing tougher rules from day one, and even established states are tightening specific standards rather than relaxing them.

When the System Fails: New York and Michigan Lab Fraud

Rules on paper only matter if labs actually follow them, and the past year gave the industry three object lessons in what happens when they don't. The most serious recent case unfolded in New York, where the Office of Cannabis Management recalled products in February and March of 2026 after discovering that Keystone State Testing New York had falsified results on 54 lots of Aspergillus testing, along with the cadmium result on one additional lot. This wasn't a clerical error repeated 54 times -- investigators found none of the affected lots actually met the state's safety standards, despite paperwork saying otherwise.

OCM Executive Deputy Director Stephen Geskey didn't mince words about what that meant: inaccurate results mean product safety simply cannot be guaranteed, and that's not a risk the office is willing to take on behalf of consumers. OCM fined the lab up to $2 million and suspended its accreditation, effectively removing it from the market. For every product that passed through that lab's doors, the result attached to it is now worthless -- and for the cultivators and processors who paid for that testing in good faith, it means scrambling to re-test inventory and explain the gap to retail partners.

Michigan's enforcement action in August 2025 described a different flavor of the same problem: a major testing lab permanently banned after regulators found a sustained, deliberate pattern of mold and potency fraud. The word sustained matters here -- this wasn't a one-time lapse but an ongoing practice that gave the lab's clients inflated potency numbers and clean mold results they hadn't earned, while honest competitors who reported accurate numbers lost business to a lab willing to cut corners. That's the mechanism by which one bad actor can distort an entire regional market.

California isn't immune either. The state shut down California Cannabis Testing Labs in 2024 after finding inflated potency results alongside missed reporting on pesticides and mold -- in one of the states with the most comprehensive testing requirements on paper. The lesson across all three cases is the same: lab shopping and deliberate fraud aren't rare, isolated incidents confined to lightly regulated states. They're a recurring failure mode that shows up even where the rulebook looks strong, because the rulebook only works if someone is actually enforcing it and auditing the labs that implement it.

THC Inflation: The Number Everyone Argues About

Nearly half (48%) of cannabis products tested in 2026 deviated more than 20% from their labeled THC content, while a Colorado study found average THC levels ran 23.1% below the lowest labeled value, highlighting widespread inaccuracies in cannabis potency labeling.

Of all the numbers on a cannabis label, none gets fought over like THC percentage -- and the data increasingly suggests that fight is justified. A 2026 study found that 48% of licensed cannabis products deviated by more than 20% from their labeled THC concentration, meaning close to half the products tested didn't come reasonably close to matching what the package said. That's not noise from natural plant variability; a 20% deviation threshold is wide enough to absorb normal batch differences and still catch something systemic.

The Colorado data tells a more specific story. A peer-reviewed study published in PLOS One, analyzing dispensary samples across the state, found average observed THC potency running 23.1% lower than the lowest value stated on the label -- not the average label value, the lowest end of the stated range. That's a meaningful gap in a state widely regarded as having mature, well-established cannabis regulation, which suggests THC inflation isn't just a symptom of young or under-resourced markets.

Regulators are starting to respond with direct intervention rather than relying on labs to self-correct. Massachusetts announced in May 2026 that it would independently re-test retail THC potency and pull any product falling outside a 75%-125% band around its labeled amount -- essentially building a state-run audit layer on top of the existing lab system. Maryland's Cannabis Administration is taking a related but distinct approach, launching a cross-testing pilot with Cannabis Safety & Quality in October 2026 aimed specifically at catching THC inflation tied to lab shopping, where cultivators route product to whichever lab reports the most flattering number.

The reason this keeps happening isn't mysterious. Higher advertised THC drives sales -- buyers routinely treat potency percentage as a proxy for quality, even though terpene content and cannabinoid ratios matter at least as much to the actual experience. That consumer behavior creates direct financial pressure on labs, which compete for cultivator business partly on the basis of who reports the most generous numbers. It's a structural incentive problem baked into how the testing market is organized, not simply a story about a few dishonest lab techs. For growers, the practical takeaway is blunt: a strain's reported THC percentage is only as trustworthy as the lab and the state regulatory regime standing behind it, and in a near-even number of cases right now, that trust isn't earned.

What This Means for Growers and Buyers

If you're growing or distributing across more than one state, the single biggest mistake is assuming a compliance plan built for one jurisdiction will translate to another. It won't. A panel that satisfies Colorado's ISO 17025 lab requirement and Massachusetts' potency-pesticide-microbial-mycotoxin combination may still fall short of what California's Department of Cannabis Control expects from a Department of Food and Agriculture-licensed lab. Treat testing compliance as something you build state by state, with its own checklist, not as a single policy you copy and paste.

Before you trust any lab's numbers, ask who accredited it. ISO 17025 is the baseline that serious states require, and a lab that can't produce current accreditation documentation -- or that's vague about which body issued it -- deserves scrutiny before you hand over product. It's also worth a quick check of recent state enforcement actions; the New York, Michigan, and California cases above are a reminder that accreditation can be suspended or revoked, and a lab's standing can change between harvests.

Apply the same skepticism in both directions. A suspiciously low pesticide result and a suspiciously high THC number are the same red flag wearing different clothes -- both can be signs of a lab competing for your business by reporting what clients want to see rather than what's actually in the sample. If a number looks too good relative to your own experience with a strain or your knowledge of the cultivar's typical range, it's worth a second opinion, not a celebration.

None of this testing infrastructure fixes bad genetics, and no lab report can manufacture consistency that wasn't there to begin with. Starting from stable, well-bred seed stock reduces the batch-to-batch variability that makes inconsistent test results hard to explain -- when a strain's cannabinoid profile is genetically consistent, a surprising lab result is much easier to flag as a testing problem rather than a growing problem. It's part of why Seedtiva breeds its seed stock with that kind of consistency as a design goal, not an afterthought. But genetics and lab accuracy are only two pieces of the equation. Your climate control, feeding program, and post-harvest drying and curing still determine most of what ends up in that jar, and no amount of good breeding or accurate testing will substitute for disciplined cultivation practice.

Nothing about this patchwork suggests it's heading toward simplicity anytime soon. If anything, it's about to get more layered: the DEA's 2026 Schedule III framework introduces a federal testing category that will sit alongside, not replace, state rules, and states like Minnesota and New York keep revising their own standards in response to specific failures they've uncovered. Expect more jurisdictions to tighten pathogen and contaminant rules over the next few years, and expect the gap between the strictest and loosest states to persist even as the floor rises overall.

What's striking about every scandal detailed here -- Keystone State Testing in New York, the banned lab in Michigan, California Cannabis Testing Labs -- is that none of them trace back to a shortage of scientific capability. The assays to detect Aspergillus, verify THC content accurately, or screen for heavy metals are well established and widely available. What broke in each case was the financial relationship between labs and the cultivators paying for results, where a lab willing to report flattering numbers can out-compete one that reports accurate ones. That's a market structure problem, and it will keep producing the same kind of fraud until the incentives change or enforcement gets sharper.

Given all that, the smartest move for both growers and buyers is to stop treating a lab report as a final answer. It's a starting point -- one data point to weigh against the lab's accreditation history, the state's specific requirements, your own experience with a cultivar, and, where it's available, a second test. The people who get burned in this industry are usually the ones who stopped asking questions the moment they saw a number they liked.