Mould on a CBD flower: what research really allows us to say

A review published in 2023 in Frontiers in Microbiology counts more than a hundred fungal species identified on cannabis and hemp flowers. Among the genera capable of producing toxins are Aspergillus, Penicillium, Fusarium, Mucor, Alternaria, Cladosporium and Botrytis.

They do not all pose the same problem. Botrytis cinerea, responsible for bud rot, shows itself through browning and decomposition of the bracts, up to the destruction of the entire bud. Penicillium, for its part, is the toxin-producing genus most frequently isolated in batch-analysis work.

This is the most useful parameter, and the least known to the general public. The same 2023 review sets out the following rule: when drying is sufficient, the water activity of the tissues drops below 0.7, which prevents the growth of most fungi.

An important nuance goes with it: Penicillium and Aspergillus can survive between 0.62 and 0.7, a range described as unfavourable to their growth without ruling it out. Drying is therefore not an absolute guarantee, but it remains the decisive lever. Work published in 2019 in Frontiers in Plant Science also shows that increasing air circulation in the final weeks before harvest significantly reduces the number of yeasts and moulds in the inflorescences.

This threshold speaks directly to storage benchmarks. The ASTM D8197 standard sets a water activity of 0.55 to 0.65 for dried flower: the upper bound is aimed precisely at staying below the level at which moulds develop. We set out these benchmarks in our article on storing CBD flowers.

Many guides suggest telling white mould apart from trichomes "by eye". We have found no scientific source validating that method for a consumer.

What the literature describes is more nuanced. Powdery mildew does produce a white deposit made up of mycelium and spores, morphologically distinct from trichomes, but reliable identification requires a microscope or a molecular analysis. As for Botrytis, the 2019 review notes that a grey-brown mycelium appears only after forty-eight hours of incubating the tissues in a sealed bag, which is nothing like a countertop test.

The honest formulation is therefore this: browning, decomposition or a fermented smell are warning signals sufficient not to consume a product, but the absence of a visible sign proves nothing. Only batch analysis documents the fungal load.

This is the strongest argument in favour of a tested product. The 2023 review reports that ochratoxin A was detected in a third of one hundred and forty-two samples of illicit cannabis, whereas it was not detected in more than nine thousand samples from legal, controlled markets.

The health stake is not theoretical. In its 2020 opinion, the EFSA scientific panel on contaminants concluded that the tolerable weekly intake of 120 ng per kilogram of body weight, established in 2006, was no longer valid, the calculated margins of exposure proving lower than the safety level adopted for almost every scenario.

The same review also reports that cannabis users showed a risk of fungal infection 3.5 times higher than that of non-users in an American insurance study, Aspergillus being the most frequent organism in immunocompromised patients. That figure is secondary within the review and deserves to be read as such.

A seriously documented batch shows, beyond the cannabinoid content, testing for microbiological contaminants. We explain how to read these documents in our article on the certificate of analysis, and the checks to run beforehand in our guide to buying quality CBD.

One contextual note, finally, specific to the French framework. On 21 March 2025, the Anses proposed classifying CBD as presumed toxic for human reproduction, in category 1B, on the basis of animal studies. That proposal, submitted to European public consultation, does not concern moulds, but it is a reminder that the question of risk does not come down to fungal contamination alone. The criteria of appearance and freshness are set out in our article on the best CBD flowers.

How do I recognise a mouldy CBD flower?

Browning of the bracts, decomposition of the tissues or a fermented smell are signals sufficient to set a product aside. By contrast, no validated visual method allows a white fungal deposit to be reliably told apart from trichomes: that distinction requires a microscope or a laboratory analysis. The absence of a visible sign therefore guarantees nothing.

At what humidity level does a flower go mouldy?

The literature places the threshold at the level of water activity, not the relative humidity of the room. Below 0.7, the growth of most fungi is prevented. Penicillium and Aspergillus can nevertheless survive between 0.62 and 0.7, without developing favourably there. The ASTM D8197 standard recommends a range of 0.55 to 0.65 for dried flower.

Are mycotoxins tested for in products sold in France?

There is no dedicated Anses or EFSA opinion on mycotoxins in CBD flowers. What is documented is the gap between channels: ochratoxin A was found in a third of a sample of illicit products, against no detection across more than nine thousand samples from controlled markets. Ask for the batch certificate of analysis and check that it does cover microbiological contaminants.