Reading DCC’s Recall and Enforcement Pages, and the Regulations Underneath Them

by

A warning sign with a cannabis leaf and exclamation point is labeled "CA." A large red stamp with the word "RECALLED" in green covers the sign, highlighting its status as public record information. Ca NORML

by Brandon Moore, Phenominal Consulting

The Department of Cannabis Control publishes every recall and every enforcement action it issues on two public web pages.1 No login, no records request, no fee. Anyone can read them, search them by license number, and walk the entire history. Almost nobody in the industry does.

That is the strange part. For the operator named in a recall, those pages are the worst day of the quarter. For everyone else, the same pages are the regulator quietly telling you, in advance, where it has been looking and what it has been finding. That is useful to anyone with a partner to vet, which is to say almost everyone in the chain: the distributor deciding whose product to carry, the brand vetting a co-packer, the cultivator or retailer weighing the distributor who will take their product to market. The cited party reads the record too late. There is nothing stopping the rest of the industry from reading it early, except that hardly anyone thinks to.

So read it. Here is what falls out when you do.

The first thing is that recall risk is not spread evenly, and it is not where most people brace for it. Of California’s roughly 7,630 active licenses, only about 140 have been named in a recall since the start of 2024.2 That is under 2% of the active universe; recalls are rare. But within that small group, the concentration is extreme: 86% of the recalled licenses are distributors or manufacturers. Count microbusinesses, which carry distribution privileges, and the share climbs past 98%. Cultivators and retailers together account for under 2% of the entire recall record.

That is not the regulator picking on distributors. It is structural, and it follows the supply chain. A cultivator sits upstream of the Certificate of Analysis. By the time a lab result becomes a printed label and that label reaches a shelf, the flower has already left the cultivator’s hands and passed through the distributor’s quality-assurance hold. That hold is the statutory checkpoint where a batch is supposed to be verified against its COA before it moves, and it is not a courtesy. California Code of Regulations Title 4, section 15307,3 puts the quality-assurance review on the distributor. Before any product moves to a retailer, the section requires the distributor to confirm that the label is “consistent with the certificate of analysis… regarding cannabinoid content,” with the labeled amounts “accurate” against the COA. The recall lands at the point of release, not the point of origin. In California’s chain, that release happens at the distributor’s quality-assurance review and on the manufacturer’s packaging line. So the recall record is, in effect, a map of where the regulated act of releasing a product happens. A distributor who has never read it is the one party in the chain least informed about its own exposure.

The second thing the record shows is a single failure that repeats, and it is not the one the language suggests.

About a third of California cannabis recalls since the start of 2024 are for a label or data failure rather than a contaminant. Inside that group, one phrase appears over and over: Inaccurate Labeling (Cannabinoid inflation). It is cited in 21 separate recalls, roughly 40% of every label-and-data recall in the record. The phrase reads like an accusation. It sounds like someone in a back office was padding a potency number to move more product.

The record points at something quieter. For dried flower, the figure the label is required to carry is Total THC, expressed as a percentage. What the citation describes is a different number landing in that spot: the value for Total Cannabinoids, printed where Total THC is supposed to be. Total Cannabinoids is the bigger figure. It is THC plus everything else the panel detected, so when it lands in the Total THC field the label overstates potency. The two numbers come from the same lab panel, which is why the pattern reads less like a lab error than like a transcription slip, something that happens between the COA and the label artwork. The recall, in any case, lands on the licensee whose name is on the package. The regulation names the figure that has to be right: section 174074 requires Total THC as a percentage on the label of nonmanufactured cannabis goods, and section 15307 makes it the distributor’s job to confirm that labeled figure is accurate against the COA before the product moves. The mismatch lives in exactly that gap, the one the regulation assigns the distributor to close.

The check that would catch it is simple: Total THC on the label has to equal Total THC on the COA. Not the bigger number. Where that check belongs, whether in software, in a packaging proof, or in a person’s hands, depends on how an operation is built. But the thing to look for never changes, and knowing what to look for is most of the job.

This is not a pattern you have to take on faith. Here are the 21 recalls where the phrase has been cited, in date order, each linked to its entry on the DCC recall page.5 Open a handful and read what the record says across different operators, different brands, 18 months apart:

Those 21 are not 21 unrelated mistakes. They are one workflow gap, repeating, in plain view, on a page the regulator has been publishing the whole time.

Which brings up the third thing, the one underneath the other two: the record only works if you can read it, and reading it means learning a vocabulary. The DCC writes its recalls and citations in a small, controlled set of phrases, and they are not interchangeable, however much they sound like it. The phrases are not evenly used, either. Counted across the recall record, the distribution is steep:6

Citation phrase Times cited What it points at
Misbranded 69 The broad bucket: any labeling noncompliance, from a missing required element to a packaging violation
Inaccurate Labeling 23 A specific data problem on the label, not a formatting one
Cannabinoid inflation 21 The COA-to-label number swap described above; almost always cited as a form of inaccurate labeling
Attractive to Children 18 Packaging or marketing that could appeal to a minor; triggers a recall and packaging-and-marketing review, separate from any potency question
Inaccurate Ingredient 4 An ingredient or formulation discrepancy
Invalid Batch Testing 2 A testing-validity failure upstream of the label

The shape of that table is the point. Misbranded dominates because it is the catch-all, cited 3 times more than anything else. Inaccurate Labeling and the Cannabinoid inflation cases nested inside it are the next tier. Attractive to Children sits just below, and it is a different track entirely: it has nothing to do with whether the printed numbers are right, and a reader who lumps it in with the labeling phrases will miss what it actually triggers, which is scrutiny of how a product is packaged and sold, not how it is tested.

The distinction most people miss is the first two. Misbranded and Inaccurate Labeling sound like synonyms and are not, and a reader who treats them as the same phrase will misread half the record. The vocabulary is small enough to learn in an afternoon, and learning it is the difference between skimming a list of other people’s problems and seeing the shape of what the regulator keeps finding.

The record carries one more signal, in when the actions land rather than what they say. Recalls and enforcement actions are not spread evenly across the license year. They cluster late in the license year: a license is about 50% more likely to draw a recall or enforcement action in the 60 days before its renewal than in the 60 days after.7 Some of that is the calendar, since California’s annual licenses concentrate their renewals in late spring and summer. But the direction holds. When you read a partner’s history, the weeks right before their renewal are where the record is most likely to have something to say.

The pages themselves are plain to use once you know what you are after. The recall page is at recalls.cannabis.ca.gov; the enforcement page, the second public record, is at cannabis.ca.gov under the department’s enforcement actions, carrying over 800 actions.8 Neither comes with a guide, and neither is built to be browsed for insight, but both let you search by license number, which is the search that matters: put in a partner’s license and read its history straight off the page. The two pages together are not a list of other people’s bad luck. They are a standing record of what the regulator checks and what it keeps finding, published in advance, free, for anyone who thinks to read it. The operators in the recall list above likely did not. The only thing separating them from everyone who comes after is that the record was already there, and they learned what it said the slow way.

You can read it the other way. A distributor can check a brand before carrying it. A brand can check a co-packer before handing over a formulation. A cultivator or a retailer can check the distributor about to take their product to market, which is the partner whose QA hold decides whether their name ends up in the record alongside someone else’s mistake. The check runs the same in every direction: pull the license number, read its history, decide with the record in front of you instead of after it.

The commercial case for spending the half hour is obvious enough. The one underneath it is simpler. Every recall in those pages is a product that reached a shelf with something wrong on it, and a consumer who bought it on the strength of a number the label could not back up. The record is the DCC’s best public account of how that happens. Reading it before any major purchase is, in the end, the cheapest thing any operator can do. It keeps their own product out of the record, while protecting the people who buy that product.

Sources and notes

  1. California Department of Cannabis Control, public records. Recalls: recalls.cannabis.ca.gov. Enforcement actions: cannabis.ca.gov (Enforcement Actions).
  2. Recall and license figures in this article are the author’s own counts, compiled from the DCC public recall record and the department’s license data, current as of June 2026. Because the DCC records update continuously, exact counts will drift over time; the figures here are rounded and reflect the record as it stood when this was written.
  3. Cal. Code Regs. tit. 4, § 15307 (Quality-Assurance Review): law.cornell.edu/regulations/california/4-CCR-15307. Subsection (c) requires the distributor to ensure the label “is consistent with the certificate of analysis for regulatory compliance testing regarding cannabinoid content,” with labeled amounts “accurate,” before transporting cannabis goods to a retailer.
  4. Cal. Code Regs. tit. 4, § 17407 (Cannabinoid Content Labeling): law.cornell.edu/regulations/california/4-CCR-17407. Subsection (b)(3) requires that nonmanufactured cannabis goods be labeled with “Total THC content expressed as a percentage.”
  5. The 21 recalls listed above, each citing “Inaccurate Labeling (Cannabinoid inflation),” with their entries on the DCC recall page at recalls.cannabis.ca.gov. Compiled by the author from the public recall record, current as of June 2026.
  6. Citation-phrase counts compiled by the author from the DCC public recall record, current as of June 2026. Counts reflect how often each phrase appears across the recalls in the record and will change as new recalls are published.
  7. Timing observation compiled by the author from the DCC public recall and enforcement records together with the department’s license-expiration data, current as of June 2026. For each action, the gap between the action date and the license’s next annual renewal was measured across the full record. In the 60 days before renewal there were about 50% more actions than in the 60 days after (roughly 170 versus 115); because every license passes through both its own before-window and after-window each year, that count ratio is also the per-license rate ratio. The same lean holds at the 90-day window. The “50%” figure compares the 60 days before renewal with the 60 days after, not with the yearly average, which is a smaller difference. California’s annual licenses concentrate their renewal dates in late spring and summer, so part of the distribution reflects the renewal calendar rather than enforcement timing; the figure is offered as an observed pattern in the record, not a causal claim about how the regulator times its actions.
  8. DCC Enforcement Actions, cannabis.ca.gov (Enforcement Actions). Action count current as of June 2026.

Brandon Moore is the owner of Phenominal Consulting, where he works with California cannabis businesses on compliance and the operational details that keep the over regulated industry viable.

Phenominal Consulting is a proud business member of California NORML. See their Cannabis Marketplace listing for more information.

Related posts:

No related posts.

Stay Informed! Join California NORML’s Email Alert List

Subscribe