Key Takeaways
- California’s updated pesticide testing regulations eliminated the outdated Category I/II system as of October 1st, introducing standardized numerical action levels.
- The new regulations require every laboratory to adhere to fixed limits for the most dangerous pesticides, preventing inconsistent testing results.
- Previously, labs set their own limits of detection, leading to ‘lab shopping’ and varying outcomes; now, all tests will be uniformly evaluated against action levels.
- The rollout of new rules occurs in two phases: Phase I sets initial action levels, while Phase II will introduce more stringent standards in 2028.
- Manufacturers must understand action levels and the significance of LOQs on Certificates of Analysis; detection alone no longer determines pass or fail outcomes.
Six years ago, we published an article explaining why California’s cannabis pesticide testing system had a serious design flaw. Labs had long exploited a quiet loophole,setting their own detection limits for the most dangerous pesticides on the market. As of October 1st, that flaw is gone.
The Department of Cannabis Control’s updated residual pesticide testing regulations took effect, and they scrap the old Category I/Category II system entirely. In its place: a single, numbers-based framework where every pesticide gets a defined action level, and every lab has to test to it. It’s not a small tweak. It’s the exact fix we argued for in 2020.
What We Reported in 2020 About California’s Pesticide Testing Problem
In our May 2020 article, “The Pesticide Problem,” we dug into how the Bureau of Cannabis Control (the DCC’s predecessor) handled Category I pesticides, a list of 21 compounds considered too dangerous to allow at any detectable level. On paper, that sounds strict. In practice, it created chaos.
The regulation at the time required a batch to fail if a Category I pesticide was “detected above the limit of detection (LOD).” The catch: each lab calculated its own LOD, based on its own equipment and methods. We gathered testing data from six different California labs and found wide swings in what counted as a detectable amount of the same pesticide.
Zach Eisenberg of Anresco Laboratories told us plainly: “Each laboratory is permitted to set their own limits of detection (LODs), so what passes at one lab might fail at another.” He also pointed out the perverse incentive this created. Labs with better instrumentation, capable of detecting smaller traces, were at a competitive disadvantage, while labs with higher LODs became more attractive to operators trying to avoid a failed batch.
We called this what it was: lab shopping. And we noted that the state had already fixed this exact issue for residual solvents by setting fixed action levels. Why not pesticides too? “Inexplicably they did not do the same for Cat I pesticides,” Eisenberg said.
How the New DCC Rule Eliminates the Old Loophole
The regulation that took effect October 1st, DCC-2025-03-R, directly answers the problem we flagged. California’s revised cannabis testing code (Section 15719) eliminates the Category I and Category II labels. Every pesticide, including the 21 that used to trigger an automatic fail on detection, now has a specific numerical action level measured in micrograms per gram (µg/g).
In Phase I, regulators assigned an action level of 0.10 µg/g to all 21 former Category I pesticides, including Aldicarb, Chlorpyrifos, and Imazalil. Labs must set their limit of quantitation (LOQ) at or below that action level for each compound. A sample now passes or fails based on a measured number, not on whether a lab’s particular equipment happened to catch a trace amount.
The DCC’s own rulemaking record confirms this was intentional. In response to public comments on the proposal, the agency explained that the changes aimed “to reduce some of the variability between laboratories” and that establishing a more uniform method of setting LODs would reduce detection variability among licensed laboratories. Other commenters went further, stating they “strongly support DCC’s effort to eliminate the practice of lab shopping, which has been detrimental to the industry and has placed honest laboratories at a disadvantage.”
That’s close to a direct nod to the exact problem we documented in 2020.
Why California Phased the Rollout Into Two Parts
The DCC didn’t flip the switch on every change at once. The rulemaking record shows the agency originally proposed much lower action levels, based on recommendations from the Department of Pesticide Regulation. But public comments and a Standard Regulatory Impact Analysis found that the initial proposal would have been too strict for current lab instrumentation and could have pushed half the state’s licensed cannabis labs out of business.
So the DCC split the overhaul into two phases. Phase I, effective today, uses action levels that labs can meet with equipment already in use. Phase II, arriving April 1, 2028, introduces 14 new pesticide analytes (including procymidone and pymetrozine) and revises a number of existing thresholds downward, some into the parts-per-billion range.
The 18-month runway between phases gives labs the time they need to develop, validate, and gain accreditation on new testing methods before the tougher thresholds kick in. The DCC acknowledged in its own responses to industry comments that building and validating a new method can take far longer than regulators initially assumed, and adjusted the timeline accordingly.
What Changes for Operators Reading a COA Today
Before today, if a Certificate of Analysis detected even a trace amount of a Category I pesticide, the product automatically failed, no matter how small that amount was. Under the new framework, detection alone no longer tells the full story.
Now, three numbers matter on every COA:
- LOD (Limit of Detection): Whether the lab can reliably tell a compound is present at all
- LOQ (Limit of Quantitation): The smallest amount the lab can reliably measure
- Action level: The regulatory threshold a result is compared against to determine pass or fail
A pesticide showing up at a trace level no longer automatically sinks a batch. What matters is whether that measured concentration exceeds the action level set in the regulation. Same chemical, same lab finding, but a very different outcome than it would have produced yesterday.
This matters most for manufacturers working with concentrates. A pesticide residue that sits comfortably under the radar in biomass can become more concentrated after extraction, which is exactly the scenario that used to blindside operators with surprise Category I failures. The new rule doesn’t eliminate that risk, but it measures it against a defined number rather than a lab-specific guess.
What This Means for the Future of Cannabis Testing in California
We’ve spent years pointing out where California’s testing system left the door open for inconsistency, and this is one of those rare moments where the fix matches the ask almost word for word. Standardized action levels ensure that regulators judge samples from Sacramento and Los Angeles by the same yardstick.
It’s not a perfect system, and Phase II will bring its own adjustment period when it lands in 2028. New analytes mean new validation work, and some of those compounds, like procymidone, aren’t even legal for use in the U.S. but still turn up in testing data from both licensed and unlicensed grows. That’s a conversation for operators to start having with their labs well before the 2028 deadline arrives.
For now, the lesson for brands and manufacturers is simple: know your action levels, ask your lab how their LOQs line up with the new tables, and stop treating “detected” as the only word that matters on a COA. The number next to it is the one that counts now.
Frequently Asked Questions
The regulation, DCC-2025-03-R, took effect October 1, 2026. It was approved and filed with the California Secretary of State on July 27, 2026.
Category I pesticides automatically failed a batch if detected at all, with the detection threshold set independently by each lab. Category II pesticides were measured against state-set numerical action levels. The new regulation eliminates this distinction entirely, giving every pesticide a defined action level.
Starting April 1, 2028, the DCC will add 14 new pesticide analytes to the required testing panel, including procymidone and pymetrozine, and will revise action levels and limit of quantitation requirements for many existing compounds.
Operators should talk directly with their testing lab about how pesticide reporting will change, review historical results for former Category I pesticides against the new action levels, and avoid making production decisions based on the old detection-only standard that no longer applies.
