Why Your COA's THC Number Won't Match the Jar Six Months Later

September 4, 2026

Photo by Jeff W via Unsplash.

Somebody sends you a photo of a COA every time they want to prove their weed is stronger than yours. Fine — but that piece of paper is a snapshot, taken at one moment, on one sample, usually pulled fresh off a drying rack or right after a final cure. It's not a promise about what's sitting in that jar three months later on a dispensary shelf or in your grow room. Nobody hands you a COA at the moment you actually smoke the product, and that gap matters more than most buyers and even a lot of growers realize.

Here's the part that doesn't get said enough: THC doesn't just sit there waiting to be smoked. It reacts. Delta-9 THC oxidizes into CBN, a much less psychoactive cannabinoid, and that conversion starts the moment a plant is harvested and never fully stops. Heat speeds it up, light speeds it up, oxygen exposure speeds it up — but even in a cool dark jar, the clock is running. A high THC number tested today is not the same chemical reality in a bag six months from now.

This is landing at an especially bad time for label trust. Through 2026, potency-inflation enforcement has hit California, Massachusetts, and Washington hard, with labs shut down, sued, or stripped of accreditation over inflated results. So the starting number is already suspect in a lot of cases — before degradation even gets a chance to quietly erode it further.

The Chemistry: How THC Becomes CBN

Photo by Jack H via Unsplash.

Fresh cannabis flower doesn't actually contain much delta-9 THC. What it contains, packed into the trichome heads on the bud surface, is mostly THCA — tetrahydrocannabinolic acid, the non-intoxicating precursor. THCA needs heat to decarboxylate, shedding a carboxyl group and converting to the delta-9 THC that actually gets you high. That's why a raw bud won't do much for you until it's smoked, vaped, or otherwise heated — combustion finishes a conversion that started slowly on its own the moment the plant was cut.

But decarboxylation isn't the only reaction happening in stored flower. Over weeks and months, oxidation takes over as the dominant degradation pathway. THC molecules react with ambient oxygen and slowly rearrange into CBN — cannabinol, a compound that's roughly one-tenth as psychoactive as THC and carries a reputation, deserved or not, for sedative, couch-lock effects rather than the clear-headed high THC delivers. A batch that's oxidized significantly doesn't just test lower on paper; it smokes differently. The high gets heavier, duller, and less predictable than what the original COA implied.

Surface area is the multiplier here. Grinding flower before storage — the way pre-ground shake, small-batch pre-rolls, or bulk trim products often get handled — exposes vastly more resin gland surface to oxygen than an intact, unbroken bud does. That's why ground material degrades noticeably faster than whole flower stored under identical conditions. If you've ever noticed a pre-ground jar losing punch faster than the same strain in bud form, this is why: more exposed surface, more oxidation, more CBN.

Curing mistakes compound the problem. Flower that goes into jars too wet, or gets burped incorrectly, traps moisture that encourages microbial activity and accelerates cannabinoid breakdown alongside oxidation. Light exposure adds another layer — UV and even visible light drive photodegradation of cannabinoids, which is exactly why clear glass jars and clear mylar bags, however nice they look on a dispensary shelf, are actively working against potency retention. Opaque, airtight, properly burped storage isn't cosmetic advice. It's the difference between a jar that tests close to its original COA six months later and one that doesn't.

What a 24-Month Stability Trial Actually Found

After 24 months of storage at 25°C, unstable cannabinoids CBGA and THCA degrade sharply (down 55% and 44%), while degradation byproducts spike dramatically—Delta-9 THC doubles and CBN skyrockets 27-fold—highlighting how cannabis potency and composition shift markedly without proper cold storage.

A stability trial published in PMC gives the clearest long-view picture of what actually happens to stored cannabis over time, and it's worth taking seriously because it wasn't a quick shelf-life check — it tracked ground cannabis material stored in vape cartridges at a controlled 25°C for a full 24 months, testing cannabinoid content at intervals along the way.

The precursor acids collapsed hard. CBGA, the cannabigerolic acid that acts as the biosynthetic parent to other cannabinoids, dropped 55% from baseline. THCA fell 44%. Total THC — the standard calculation combining THCA and delta-9 THC — declined 21% over those two years. None of that should surprise anyone who understands oxidation kinetics, but the magnitude is bigger than most casual assumptions about how stable cannabis is in storage.

The more counterintuitive finding is what happened to the individual compounds within that total. CBN increased by 2700% — not a typo, nearly a 28-fold jump — as THC oxidized into it over the two-year window. Delta-9 THC itself actually increased 100%, doubling from its starting point, because THCA in the material kept slowly decarboxylating into active THC even in the sealed, non-combusted storage state, before that fresh THC itself began oxidizing onward into CBN. It's a moving chemical system, not a fixed number sitting quietly on a lab report.

Temperature was the clearest lever in the whole dataset. Samples stored at 30°C showed steeper, faster shifts across every one of these compounds than the 25°C samples, confirming that heat — more than time alone — drives the pace of degradation. Five degrees made a measurable difference over two years, which says a lot about what a hot delivery truck or an un-air-conditioned stockroom does to product on a much shorter timeline.

Put together, this trial says something uncomfortable but important: a COA generated at harvest or at packaging isn't describing a stable endpoint. It's describing one frame of a slow-motion chemical reaction that keeps running under the consumer's own roof, in their own jar, at whatever temperature their house happens to sit at.

Temperature, Light, and the First 30 Days

Anresco Laboratories ran a storage study specifically isolating temperature and light as variables, and the temperature numbers alone make the case. Samples held at a refrigerated 4°C retained 3.2% more total potential THC than identical samples held at 20°C — roughly typical room temperature — and 14.1% more than samples held at 30°C, a temperature not unusual for a parked delivery van in July or a poorly ventilated stockroom.

What stands out in their data is that thermal degradation outpaced photodegradation by a wide margin. Light exposure mattered, but heat did more damage, faster. That reframes a lot of conventional advice that focuses mostly on keeping product out of sunlight while saying less about temperature control. A dark cabinet that happens to sit near a heat vent or in a warm garage is doing less good than people assume.

The timing detail matters most for anyone trying to connect this back to the label on the package. The steepest drop in potency across all storage conditions in the study happened in the first 30 days. Not month six, not month twelve — the first month. And the first 30 days after compliance testing is very often the exact window a product spends moving from cultivator to processor to distributor to dispensary shelf before it ever reaches a customer's hands. The COA gets printed near the start of that window. The steepest decay happens during it.

Practically, this argues for treating storage conditions as seriously as any other part of quality control. Glass jars over plastic bags, humidity control packs to keep RH stable without introducing excess moisture, cool storage — refrigeration for anything held long-term in bulk — and minimizing how long product sits between test and sale all meaningfully slow the clock down, even though nothing stops it entirely.

Home growers curing their own harvest should read this as a direct operational note: the first month post-harvest, through cure and into early storage, is the single highest-stakes window for holding onto whatever potency a test — home or lab — actually measured. Get sloppy with jars, humidity, or temperature in that window and you're giving away potency faster than at almost any later point in the product's life.

The Regulatory Blind Spot: One Test, No Expiration Date

Photo by Budding . via Unsplash.

Here's the regulatory gap that makes all of this worse: shelf-life testing simply isn't required by most U.S. cannabis regulatory bodies right now. A COA is a pass/fail compliance snapshot tied to one batch at one point in time — it certifies that a sample met potency and safety thresholds on the day it was tested, and it says nothing, contractually or scientifically, about the weeks or months that follow.

California's Department of Cannabis Control allows a label variance of 12% for edibles and 10% for flower and other manufactured goods — meaning a labeled THC percentage can legally differ from the lab-measured result by that margin and still pass compliance. But that tolerance is anchored entirely to the static compliance-batch test point. It has nothing to do with degradation. A product could be perfectly within variance the day it's tested and meaningfully outside any reasonable expectation of that number by the time a customer buys it weeks later — and no rule currently on the books requires anyone to account for that gap.

This is a separate problem from the industry's broader potency-inflation issue, but the two compound each other badly. One is a testing-integrity problem — labs reporting numbers that don't reflect the actual sample. The other is a physics problem — real cannabinoids genuinely changing over time regardless of how honest the original test was. Stack them together and the number on a label can be wrong for two completely independent reasons simultaneously.

The 2026 enforcement wave is a direct response to the first problem, and it's been aggressive. California lost 27% of its licensed testing labs in a single year once state auditing tightened and inflated results got flagged. In Massachusetts, MCR Labs is suing eight competing labs over allegedly inflated potency reporting. Washington regulators shut down testing lab Praxis entirely. None of these actions touch the degradation issue at all — they're purely about labs gaming the initial number. Which means even in a best-case world where every lab reports honestly, the underlying degradation blind spot in the regulatory framework remains completely untouched.

Why Nearly Half of Dispensary Labels Are Already Wrong

Photo via Pexels.

Layer potency inflation and natural degradation on top of each other and you get a genuinely troubling picture for anyone buying off a dispensary shelf. An inflated starting number doesn't just start high — it degrades downward from a baseline that was already fictional, meaning the real gap between label and lived experience compounds rather than simply adding.

Industry data on this is blunt: nearly half of licensed cannabis products sitting on dispensary shelves carry THC labels wrong by more than 20%. That's not a rounding error or a testing-variance technicality — that's a coin-flip's worth of products where the number printed on the package meaningfully misrepresents what's actually inside, before degradation even gets factored in.

One consumer-facing study measured this gap directly and found average measured THC coming in more than 23% lower than the lowest reported value on the package label — not the average claimed value, the lowest one on the range. That's the kind of finding that should reframe how anyone reads a percentage printed in bold on a jar.

Regulators are starting to respond structurally rather than just case-by-case. California and Colorado are both rolling out new tracking requirements through 2026, frequently triggered by potency audits where state agencies pull product directly off dispensary shelves for independent retesting rather than relying solely on producer-submitted compliance samples. That shift — from trusting the initial test to actively re-verifying shelf product — is a tacit admission that the current system produces numbers that don't hold up.

For a buyer, the honest takeaway isn't complicated, even if it's uncomfortable: the number on the label was likely optimistic the day it was printed, for reasons having nothing to do with chemistry, and it's only gotten less accurate since, for reasons that have everything to do with chemistry. Both problems point the same direction — toward skepticism of the printed percentage as a precise, current fact.

What This Means for Growers and Buyers

Photo by CRYSTALWEED cannabis via Unsplash.

None of this means testing is pointless — it means testing needs to be read differently than most people currently read it. For home and craft growers who do test their own harvest, the most useful practice is testing as close to point-of-use as is practical, and treating whatever number comes back as a best-case figure that starts decaying the moment it's recorded, not a fixed spec that holds indefinitely.

The single biggest lever any grower actually controls here is cure and storage, and it's worth treating with the same discipline as a nutrient schedule or a light intensity target. That means holding relative humidity in the 58-62% range through cure using calibrated hygrometers and humidity packs, storing in opaque containers rather than clear glass or mylar, keeping storage temperature cool and stable, and holding off on grinding flower until it's actually about to be used rather than pre-grinding for convenience. Every one of those choices directly slows the oxidation curve described above — none of them are cosmetic.

Genetics set the ceiling before any of this even applies. A plant grown from quality, well-bred seeds — the kind of stable, well-tested genetics Seedtiva works with — starts with a stronger, more consistent cannabinoid baseline, which gives a grower more room to work with even as normal degradation does its slow, steady work over time. Weak or unstable genetics simply have less potency to lose in the first place.

Buyers need a similar mental shift: read a COA as a historical document describing one lab's finding on one specific day, not a live spec sheet describing what's currently in the package. That reframing alone changes purchasing behavior for the better — weighing a trusted cultivator or trusted source of flower more heavily than chasing whichever label prints the highest number, since that highest number is exactly the one most likely to be inflated, most rapidly decaying, or both.

And it's worth saying plainly: there's no universal fixed decay rate here. Outcomes vary by genetics, by climate, by cure technique, by storage setup — the 24-month and 30-day figures above describe general oxidation curves under specific controlled conditions, not a guarantee that applies identically to every jar in every climate.

A COA tells you what a lab measured, on one sample, on one day, under one set of storage conditions up to that point. It was never built to track what happens after that jar gets sealed, boxed, shipped, shelved, and eventually opened at home — and expecting it to do that job is where a lot of the frustration and distrust in this industry actually comes from.

The fix isn't rejecting testing or assuming every lab is lying, even though enforcement actions suggest plenty have been. The fix is treating a potency number as a starting point on a decay curve rather than a fixed fact, and taking cure and storage conditions exactly as seriously as the rest of a cultivation process — because chemically, that's exactly what they are: an extension of the grow, not an afterthought once harvest is done.

As enforcement keeps tightening through 2026, expect more scrutiny landing on inflated initial numbers — more lab shutdowns, more lawsuits, more shelf audits. That's a genuinely good and overdue correction. But it only fixes half the problem. The oxidation curve doesn't read regulatory bulletins. THC will keep quietly turning into CBN in jars and cartridges everywhere, on its own schedule, regardless of what enforcement cleans up or what number was printed on the label to begin with.