How to read this chart, and why the numbers vary so much
The variance is the headline. The 2022 Environmental Health Perspectives review by Jameson and colleagues found that different jurisdictions showed significant variations in regulated contaminants and action levels ranging up to four orders of magnitude. There is no federal standard to anchor to, because, as the same paper notes, cannabis is neither federally regulated as an agricultural, food, nor pharmaceutical commodity. State rules fill the gap, and they fill it inconsistently.
Two cautions before the tiers. First, the number depends on the product form: a flower limit and a concentrate limit are often different within the same state, so match the limit to what you are actually testing. Second, these numbers get revised; New York, for example, loosened its yeast-and-mold rules after many of the published comparisons were written. Treat any figure here as a tier marker, not a citation, and verify the live limit with the state regulator or your lab.
The three TYM threshold tiers
States sort into three practical bands for total yeast and mold on flower.
The standard tier, and the modal one, sits at 10,000 CFU/g. As the reporting on this topic puts it, the majority of states have a 10,000 CFU threshold for total yeast and mold. Washington DC, Ohio, Vermont, and North Dakota all land at or around 10,000 CFU/g for dried flower, which makes this the number most operators design to by default.
The looser tier sits above 10,000 CFU/g. Connecticut is the widely cited outlier here; investigative reporting framed it as having a higher threshold for total yeast and mold in medical cannabis than many other states, and Maryland, Michigan, and Florida are reported among the highest. A looser TYM number does not mean a lower bar overall, because the strict states often move their enforcement to Aspergillus instead.
The stricter tier sits below 10,000 CFU/g, and a few states enforce effectively zero-tolerance microbial rules. The practical rule for a multi-state operator: design to the strictest state in your footprint, not the loosest, so one SOP clears every market. Each state's exact figure, rule citation, and enforcement history lives on its per-state page.
Aspergillus rules are separate, and they fail you differently
The most common operator error is treating a TYM limit as the whole mold rule. It is not. A separate set of states runs a pathogenic Aspergillus panel, four species, A. flavus, A. fumigatus, A. niger, and A. terreus, on a presence-or-absence basis. Michigan's rule, for instance, is that if Aspergillus spp. is detected, the sample fails, no CFU count involved.
That is why a loose TYM number and a tough compliance regime can coexist: a state can allow a high aggregate mold count and still fail your batch on a single Aspergillus colony. More than a dozen states enforce Aspergillus separately, and conflating the two panels is the top reason an operator who optimized for a good TYM number still fails a test. If your footprint includes an Aspergillus state, your prevention has to be species-aware, which is thecultivation IPM expertiselayer.
What the failure rates actually say
Stricter numbers do not map linearly to more failures, and the data shows why. Reporting on state results found that in Maryland about 5.13 percent of samples fail tests for yeast and mold, while in Michigan about 5.99 percent of tests fail. The Jameson review calculated a 2.3 percent failure rate for California flower and 9.2 percent for extracts, for a roughly 5.1 percent overall rate.
Read those together and the lesson is that Aspergillus zero-tolerance rules, not the TYM ceiling, drive most failures in states that enforce them. Michigan's elevated failure rate sits alongside its Aspergillus-detection rule, not a stringent CFU count. For an operator, that means prevention effort should follow the panel your state actually fails people on, which is often the species screen, not the aggregate count.
The federal EPA context every operator has to account for
There is no federal EPA pesticide pathway for cannabis, and that shapes what you can legally spray. The Jameson review states it directly: the U.S. Environmental Protection Agency has registered only 59 biopesticides and 1 conventional pesticide for hemp, but no pesticides for cannabis. State rules fill that vacuum, which is exactly why the thresholds vary so widely.
What it means operationally: any surface disinfectant you use in the facility must carry a valid EPA registration for its surface-use claim, and the label governs where it may go. ProKure V is a liquid, water-activated, hospital-type ClO2 disinfectant, EPA Reg. No. 87508-3-89334, used at 10 minutes at 100 ppm on hard, non-porous surfaces; it leaves no residue and is suitable for use with plants in the room, applied to surfaces rather than the crop. The empty-room air reset, ProKure G, carries the disclaimer "Not for use with plants, crops or feed" and runs only in an unoccupied room, while ProKure D is a slow-release gas for in-room air quality up to 30 days and is not an EPA-registered pesticide. Keep those lanes straight, route the air layer toair quality management, and never apply a plant-restricted product to the crop, because using an EPA-registered product inconsistent with its label is a violation of federal law.
How to use this chart across a footprint, and how to prevent the failed test
For amulti-state cannabis operator, the procurement move is to standardize on the strictest state in your footprint. One protocol tuned to the toughest market clears the rest, which beats maintaining a different SOP per facility and defending a patchwork at audit. Pair that with one sanitation chemistry across sites, varying only the documentation and the lab panel by state.
Preventing the failed test in the first place is a between-cycle discipline, not a post-failure scramble. Reset hard, non-porous surfaces with ProKure V during turnover, hold ambient air quality with ProKure D while the crop is in, and run ProKure G for the fast empty-room reset, all logged as a documentedmold prevention protocol. Procurement teams get SDS, EPA labels, and account pricing through acannabis procurementaccount, with deep label and SDS detail on prokuresolutions.com and reorders on shop.groclarity.com. GroClarity builds this cross-state standardization from its Phoenix, Arizona headquarters at 5013 E. Washington St. Ste. 130, Phoenix, AZ 85034, reachable at 866.206.1301. Built on science. Backed by results.


