Skip to main content

INSIGHTS BLOG > The Health Costs of Moldy Weed

The Health Costs of Moldy Weed

Written on 25 April 2023

Ruth Fisher, PhD. by Ruth Fisher, PhD



What is the value of using cannabis products tested for mold, that is, the avoided healthcare costs associated with not consuming moldy cannabis? We can call this value a “health premium” associated with using tested cannabis.

The point of this analysis is that cannabis products sold in licensed markets, which (theoretically) have been tested for mold and found to be mold-free, presumably have a much lower probability of being contaminated with mold than do cannabis products sold in unlicensed markets, which have not been tested. If the health premium is high enough, then the assurance that licensed products will be mold-free should make the higher prices in licensed markets worth paying.

Generally speaking, the way you normally approach such a problem is to estimate two separate factors:

  1. What’s the probability that cannabis products purchased in licensed vs. unlicensed markets is moldy?
  2. What are the health costs due to consuming moldy cannabis?

The total expected cost of consuming moldy cannabis is then the product of these two terms:

Total Expected Costs of Consuming Moldy Cannabis

= [Probability Purchased Cannabis Is Moldy]

x [Health Costs due to Consuming Moldy Cannabis]

The corresponding health premium associated with consuming tested cannabis rather than untested cannabis is then

Health Premium from Tested Cannabis

= {[Prob Untested Cannabis Is Moldy] - [Prob Tested Cannabis Is Moldy]}

X [Health Costs due to Consuming Moldy Cannabis]

In the case of moldy cannabis, it turns out we need three separate factors:

  1. Probability a purchased cannabis product is moldy (separately for tested and untested cannabis)
  2. Probability a consumer who consumes moldy cannabis gets sick
  3. Cost of treating health problems due to consuming moldy cannabis

For example, suppose:

  • There’s a 25% chance a cannabis product I buy is moldy,
  • There’s a 5% chance I get a lung infection from smoking moldy cannabis, and
  • The cost of treating a lung infection is $2,500.

In this case, the expected cost of moldy cannabis is 25% x 5% x $2,500 = $31.25. In other words, if I were a rational consumer, I would be willing to pay up to $31.25 for an assurance that the cannabis I buy is not moldy.

Suppose the price of a particular cannabis product from a licensed source is $50, while the price of the “same” product from an unlicensed source is $35. The licensed product has presumably been tested, so it’s likely the product is mold-free, but the unlicensed product can offer no such assurance. Since the value to me of that assurance is $31.25, I should be more than willing to pay the added premium ($50 - $35 = $15 < $31.25) for the licensed product.

Sounds pretty straightforward. As with everything else in cannabis, however, this analysis proved to be a lot more complicated than I had initially thought it would be.

Background Information

About Mold

A concise and well-written article by Danny Smolch, aptly titled, Cannabis and Mold: Everything There is to Know, collected most of the basics on mold together into a short read. He reports that “Mold is a group of microorganisms that grow in humid and damp surroundings with poor air ventilation” and reproduce by generating spores. Its role is to decompose dead leaves, trees, and plants. Mold spores are “exceedingly hard to get rid of” once they develop, since they can withstand heat from a lighter (combustion) or oven (decarboxylation). Mold can develop either during cultivation or storage in cold, humid environments. To prevent mold from developing, store products in dark environments with 59% - 63% humidity, at about 77°F.

Sources generally agree that healthy people who consume small amounts of moldy cannabis should feel no real adverse effects. Yet, problems may occur in certain circumstances, which are discussed in more detail later in the analysis. Regardless, it is generally advised that when moldy cannabis is detected, users should throw it out. 

Finally, while most of the discussions of moldy cannabis relate to flower, mold can also infect extracts. In fact, when cannabis is extracted, any toxins can become much more concentrated.

Licensed vs. Unlicensed Cannabis

Unlicensed (illicit) cannabis markets exist throughout the US. Over the past quarter century, most states have legalized medical cannabis, and a large number have also legalized adult use cannabis. In all legal states, licensed cannabis activity (I use the term cannabis to mean marijuana, or cannabis with ≥ 0.3% THC) is heavily regulated. Hemp or CBD activity (cannabis with  < 0.3% THC), on the other hand, occurs in separate markets from cannabis activity and is generally only lightly regulated. Heavy regulations in cannabis create high operating costs and thus high prices, so licensed market cannabis prices are generally much higher than the prices found in unlicensed markets, making it difficult for licensed sellers to compete with unlicensed sellers. The two main selling points favoring licensed over unlicensed cannabis are that licensed cannabis is (i) legal and (ii) tested. 

Most people generally want to obey the law and will tend not to engage in illegal activity, even if they have no moral qualms about the activity itself. In other words, many people will use cannabis, but only if it’s legal, and these people will only purchase cannabis in licensed markets.

As for testing, states that have established licensed cannabis markets – for either medical and/or adult use – generally require all products to be lab tested before they can be sold to consumers. (Hemp products are generally not required to be tested.) Testing provides consumers with information on product contents (e.g., the amount of THC, CBD, and other active compounds) and ensures products are safe to consume. 

Let’s dig a bit deeper into the lab testing process. Cannabis products are separated into batches by product type and size, where separate batches each require separate tests. For example, in California, flower must be divided by cultivar (strain) into batches of no more than 50 pounds. Vape cartridges, tinctures, edibles, lotions, etc. are each tested separately by concentration, formulation, and/or “flavor.” From each batch, (theoretically) a random sample is drawn and sent to an independent testing lab, where the samples are tested for potency and safety. Lab testing must generally establish (specific requirements vary by state):

  • Concentrations of active ingredients
  • Moisture content
  • Less-than-threshold amounts of toxins
    • Pesticides
    • Residual Solvents
    • Bacteria and Mold
    • Heavy Metals

Some products that fail testing requirements may be remediated and re-tested. Products that fail testing must be destroyed.

A panel of lab tests (each bullet item above requires a separate test) generally costs several hundred dollars to have done. The vast majority of unlicensed sellers do not have their products lab-tested, and savings on testing costs contribute to the ability of unlicensed sellers to undercut licensed sellers on price.

The sounds pretty straightforward, but it hasn’t been in practice. The legal cannabis industry is new, cannabis is a complex product with many active ingredients (e.g., cannabinoids and terpenes) and forms of use (e.g., flower, concentrates, tinctures, pills, lotions, edibles) that, until recently, had never been systematically tested. Furthermore, because cannabis is federally illegal, the EPA has refused to provide any guidance on either the choice or the levels of chemicals, bacteria, or fungi that should be considered toxic. As a result, testing equipment, standards, protocols, and requirements have not been standardized either within, nor certainly across, states. What this means is you can send a cannabis sample to three different labs and get back three different sets of results.

It gets worse. Testing generally has not been enforced by state cannabis commissions. Given the value of the products at stake – a batch of cannabis products may retail for tens or hundreds of thousands of dollars – there is a nontrivial incentive to fudge the results. While many sellers evade testing altogether, some other sellers pay testers to report THC potencies above the actual amounts (higher THC concentration products sell for a premium) or pass products that are contaminated by pesticides, solvents, or mold. In fact, “lab shopping”(the practice of sellers seeking labs that will report desired, rather than actual, results) is currently creating large problems for brands who are trying to play by the rules.

Finally, it’s common knowledge that much unlicensed (i.e., under-the-table) cannabis is sold through licensed dispensaries. 

A random sampling of cannabis products from licensed dispensaries that were (theoretically) tested and reported to be mold-free would therefore most likely indicate that some samples were, in fact, infected with mold.  

One final complication is that products may be tested for mold and bacteria and found to be clean, but then later become moldy or contaminated after being improperly stored (either by sellers and/or by consumers). 

Singling Out Mold

All the groups of substances included in cannabis testing requirements – pesticides, residual solvents, mold, and heavy metals – are potentially toxic when consumed. Also, “tested cannabis” generally refers to cannabis that has been tested for all four sets of toxins, that is, the four sets tend to be considered collectively, rather than individually. So then why single out mold and focus the analysis on that single group of toxins, rather than considering all four sets as a whole? 

It essentially comes down to the fact that some of the potential adverse effects of consuming mold are acute, while the effects associated with consuming the other types of toxins tend to be chronic. By focusing the analysis on mold, I can make the case for spending more money today on a single purchase so as to avoid a potential bad outcome today, and where the bad outcome can easily be attributed to today’s purchase. In other words, it's relatively quick and easy to match the cost today (the higher price paid for lab tested cannabis) to a tangible reward realized today (not getting sick from mold). The alternative is to focus on all four sets of toxins and try to make the case for spending more now to avoid an event that may occur sometime in the distant future, to which today’s purchase may or may not have contributed. That’s a really tough sell.

Price Differentials: Licensed vs. Unlicensed Cannabis

The difference in price between cannabis products purchased from licensed and unlicensed sources are both time and location specific. Licensed market prices increase with taxes and regulations, while unlicensed market prices decrease with supply. So then given the seasonal nature of (outdoor) cannabis products, price differences probably increase in the fall and winter (when the harvest comes in and unlicensed prices drop) and decrease in the spring and summer. Also, as supplies have generally increased over time throughout the country, prices have accordingly decreased. 

Cannabis flower is sold by the gram or the ounce, and a standard joint is half a gram.

As of Jan 2023, according to, the average price in the US is: 

$10/g (ranging from $9 to $15 depending on location)
$280/oz (ranging from $252 to $420 depending on location)

A Feb 2016 article by Priceonomics, Is It Cheaper to Buy Weed on the Street or at a Dispensary?, reported price differentials that were all over the map, which is expected since states vary in both their extent of regulations and natural affinity for cultivating cannabis. In some cities, unlicensed products actually sold for more than licensed market products, while in other cities there was a substantial price premium for licensed market products. Here’s a sampling of licensed market price premiums in ascending order:

Mesa, AZ:                  -8%
Denver, CO:               -3%
Vancouver, WA:           9%
Los Angeles, CA:        14%
San Diego, CA:           29%
Glendale, AZ:             35%
San Jose, CA:              45%
Sacramento, CA:         65%
Grand Rapids, MI:       68%
Medford, OR:              73%

Another article, Dispensary Prices VS Street Prices Compared 2023 Update, reports price differences between dispensary products and street products in California:

2016:               $299 dispensary vs. $218 streets (37% price premium)
2021-2022:      $450 dispensary vs. $220 streets (104% price premium)

Information for Calculating Estimates

So now back to the task of estimating 

            Health Premium for Tested Cannabis

                        = {[% Untested Cannabis Is Moldy] — [% Tested Cannabis Is Moldy]}

                                    x [Probability Consuming Moldy Cannabis Causes Illness]

                                                x [Cost of Illness]

Information Needed

Probability Cannabis Is Moldy

I need two different sets of estimates for the probability a cannabis product is moldy, one for unlicensed cannabis products and another for licensed cannabis products. And what I’m looking for in this regard is a study that says something to this effect: 

We bought 30 different cannabis products from a variety of unlicensed sources and 30 cannabis products from a variety of licensed dispensaries. We then tested them for mold and found 20 of the unlicensed (66.7%) products and 2 of the licensed products (6.7%) contained mold.

Probability Consuming Moldy Cannabis Causes Illness

The second data element I need for my calculations is the probability that if I consume moldy cannabis, I’ll get sick. Ideally, what I’m looking for is something to this effect:

We canvassed the country for cases of moldy cannabis and found a dozen instances in which batches of cannabis products were infected with mold. A total of 200 moldy products were consumed by 200 consumers, of whom 50 consumers (25%) contracted lung infections. 

Cost of Illness

The last data element, the cost of an illness, is generally the most difficult to estimate. The cost of an illness may include any or all of the following components:

  • Cost of medication
  • Cost of doctor visit
  • Cost of hospitalization
  • Cost of lost work days/productivity

When illnesses are common, then often times researchers have provided cost estimates. If the illnesses are not common or are not well-defined then the costs will be more difficult to estimate. 

Information Found

To recap, the information I am looking for is:

  • Probability cannabis is moldy (i.e., prevalence of moldy cannabis), separately for unlicensed and licensed cannabis products
  • Probability consuming moldy cannabis causes illness
  • Cost of illness due to consuming moldy cannabis

I spent a good amount of time searching for the information/studies described above, and I found five different types of relevant information:

  • Types of cannabis mold
  • Prevalence of moldy cannabis
  • Illnesses caused by moldy cannabis
  • Prevalence of fungal infection in cannabis users
  • Costs of illnesses due to mold infection

Types of Cannabis Fungi

I found three studies (there are plenty more) where researchers collected a variety of cannabis samples and tested them for contaminants:

The most commonly found molds are presented in Figure 1.

Figure 1

The most commonly mentioned forms of mold are AspergillusBotrytis (grey mold), and Golovinomyces (white mold or powdery mildew).

Prevalence of Moldy Cannabis

The next step was to find estimates of the prevalence of mold in cannabis samples, from both unlicensed and licensed sources. I found two studies, both for unlicensed sources, and both of which found that most cannabis samples from unlicensed sources were infected with some type of mold. I couldn’t find any similar studies conducted on samples of licensed cannabis. As previously mentioned, licensed cannabis products are required to be tested for mold, and in theory, this means that cannabis products sold through licensed dispensaries should generally be mold-free. However, given the problems associated with cannabis testing previously discussed, we can’t necessarily make this assumption.

For completeness sake, I’ll summarize the two studies I found on the prevalence of mold in unlicensed cannabis samples.

1982 Wisconsin Study

A 1982 study, SL Kagen et al, Marijuana smoking and fungal sensitization,  was conducted by researchers from the Medical College of Wisconsin. 

The study involved 28 “randomly selected” subjects (16F, 12M), with a range of intensity of cannabis usage, and of whom 18 (64%) were also tobacco smokers. Studies with subjects who smoke both cannabis and tobacco tend to be less valid, since outcomes may be due to tobacco use, rather than cannabis. 

Subjects provided samples of their cannabis to the researchers for testing, and 13 of 14 samples provided “contained potentially pathogenic fungi,” which included Aspergillus, Mucor, Penicillium, and thermophilic actinomycetes. The study predates legalization of cannabis, so all products are from unlicensed sources.

Of the 28 cannabis smokers:

  • 1 had systemic aspergillosis (4%) (spread of infection beyond lungs to other body parts)
  • 7 had a history of bronchospasm after smoking (25%)
  • 21 were asymptomatic (75%)

The subject with systemic aspergillosis was the second heaviest cannabis user of the 28 subjects and presumably also experienced bronchospasm after smoking. For the 7 subjects with bronchospasms, the researchers did not indicate their intensity of use nor whether or not they smoked tobacco. This makes it hard to draw any conclusions about the extent to which the bronchospasms were due to smoking (either tobacco or cannabis), inhalation of fungus, cannabis use, or tobacco use.

Precipitins (antibodies) to Aspergillus antigens (precipitins are found with repeated exposure) were found in:

  • 13 of 23 cannabis smokers (57%)
  • 1 of 10 control subjects (10%)
  • 3% - 10% found in normal populations

Both cannabis smokers and control subjects showed a similar pattern of precipitins to antigens from other fungi. It’s not clear why the test for Aspergillus antigens was not run on the 5 excluded subjects, and eliminating them from the calculation could bias estimates of the prevalence of antigens upward. If we include the full sample, we get a lower incidence of 13/28 = 46%.

Significant blastogenesis to Aspergillus was found in 3 of 23 (13%) cannabis smokers, 2 of whom were the heaviest cannabis smokers.

The conclusions I draw from this study are these: Most unlicensed cannabis may very well contain potentially pathogenic mold, yet most smokers do not exhibit symptoms from mold infection. It’s thus common for people to be exposed to mold, from cannabis or other sources, without it causing any health problems. At the same time, chronic inhalation of aspergillus-tainted cannabis may cause problems.

2016 California Study

2016 California Study, GR Thompson III et al, A microbiome assessment of medical marijuana,  was conducted by researchers at UC Davis and Steep Hill Labs. Researchers collected 20 samples of cannabis from various dispensaries (presumably licensed) in northern California and tested them for mold. They found infections of various intensities (indicated by a heat map) by 31 different species in 19 of the 20 cannabis samples. 

Note that California phased in lab testing requirements for licensed cannabis products in 2018. This study was thus conducted before these requirements had been implemented, which means there’s every reason to believe the samples were not tested.

The researchers made no indications about how many of the mold infections were intense enough to cause health effects in users. Rather, the general purpose of the study was to “educate practitioners about the potential risks” of licensed cannabis.

The conclusion I draw from this study is that most untested cannabis probably contains potentially pathogenic mold.

Illnesses Caused by Moldy Cannabis

The next step in the analysis is to determine which illnesses are caused by ingestion of moldy cannabis. Based on the fact that large numbers of people have been smoking cannabis forever, that moldy cannabis is probably very common, and that there’s very little mention of anyone becoming seriously ill after smoking moldy cannabis (certainly not after a single use), it’s probably safe to assume that smoking moldy cannabis is generally not a problem for most people. On the other hand, for people with weakened immune systems, such as those with HIV, cancer, transplants, or autoimmune diseases, smoking moldy weed can be deadly. Also, people with respiratory problems, such as asthma, may experience more serious reactions from ingesting mold.

A summary of the potential health effects from smoking moldy cannabis is provided in Figure 2.

Figure 2

2 health effects

Prevalence of Illnesses Caused by Moldy Cannabis

I found very little information about the likelihood that cannabis users who consume moldy cannabis become ill. There’s the information I mentioned in the previous section, together with the following study, which does not seem to provide any useful information.

2020 CDC Study

A 2020 study by researchers at the CDC compared the prevalence of mold infections in cannabis users to non-users by investigating health insurance claims: K Benedict et al, Cannabis Use and Fungal Infections in a Commercially Insured Population, United States, 2016.

Researchers found health insurance claims for 53,217 patients who reported using cannabis, of whom 40 (0.075%) had fungal infections, and 21,559,558 claims for patients who did not report using cannabis, of whom 6,294 (0.029%) had fungal infections (see Figure 3). Patients who reported using cannabis were thus 2.6 times more likely to have a fungal infection. Partitioning fungal infections into mold infections and other fungal infections indicates reported cannabis patients are 3.4 times more likely to have mold infections than non-cannabis patients. The analysis also found that relative to non-cannabis patients, cannabis patients are younger, 2.1 times more likely to have other fungal infections, 2.0 times more likely to be immunocompromised, and 4.5 times more likely to use tobacco.

Figure 3

3 fungus mj nomj

The researchers provide several hypotheses for their findings:

  • People who use cannabis tend to be sicker (notably, cannabis users may be using cannabis to treat their illnesses) and more immunocompromised, making them more vulnerable to fungal infections. The source of the fungal infection could not be determined, so it’s not clear it was caused by cannabis.
  • Smoking-induced lung damage may increase susceptibility to fungal infections.
  • Tobacco smoking or fungal-infected tobacco may confound or cause the results.
  • Cannabis use may have been under-reported in the health claims, which would bias ratios (column [c] in Figure 1) upward. Study data indicate 0.25% of patients use cannabis, while the US population reports about 9% use.

Limitations of the study were also noted:

  • Researchers were unable to distinguish smoking from other forms of cannabis ingestion. Non-smoking forms of use may make patients less susceptible to fungal infection.
  • Researchers identified immunocompromised status associated with diseases (HIV, cancer, transplant) but not with medications (e.g., corticosteroids and tumor necrosis factor inhibitors), which may bias the associated findings.

The conclusions I draw are the following. The study is based on an analysis of health insurance claims, which does not provide a representative sample from the larger population, but rather, represents a cross-section of less healthy people in the population. From the discussion, we know less healthy people are more likely to be affected by moldy cannabis, which biases the results. Additionally, there’s a very good chance the patients in the study under-reported their use of cannabis, perhaps by as much as 36 times, which also biases the results. I’m thus inclined to find the study unreliable and unusable for the analysis at hand.

Cost of Infections Due to Moldy Cannabis

The last set of information needed is a dollar figure for the cost of experiencing health effects associated with consuming moldy cannabis, that is, the effects listed in Figure 2. The effects of consuming moldy cannabis range from mild to severe. Perhaps the best way to approach the problem is to consider the costs of treating pneumonia, a common health problem, and thus one whose associated costs are more readily available. 

In Trends in healthcare utilization and costs associated with pneumonia in the United States during 2008–2014, Sabine Tong et al (2018) conducted a study of health insurance claims data for people with pneumonia and provided estimates of the average payment per episode during 2013 – 2014, as displayed in Figure 4.

Figure 4

4 pneumonia costs

What Figure 4 shows is that a relatively mild case of pneumonia may cost roughly $500, a moderate case about $1,400, and a severe case about $14,000. We can use these estimates as very rough proxies for the costs of seeing a doctor (outpatient visit) for a moderate respiratory infection, going to the ER for a severe infection, or being hospitalized for a dire infection.

Value Premium Model

The question this analysis seeks to answer is:

What’s the expected value of using cannabis products that have been tested for mold, rather than untested products, as measured by the expected health cost savings? 

The equation that estimates this expected health cost savings, or health premium, is:

[EQ1]   Health Premium for Tested Cannabis

[EQ2]               = {[% Untested Cannabis Is Moldy] — [% Tested Cannabis Is Moldy]}

[EQ3]                           x [Probability Consuming Moldy Cannabis Causes Illness]

[EQ4]                                       x [Cost of Illness]

I created a model (see Figure 5) that incorporates the components EQ2, EQ3, and EQ4 and that uses estimates guided by the information I’ve gathered and described thus far. 

Figure 5A: Cannabis Mold Testing Value Premium Model

 5a model 2

Figure 5B

5b model graph 2

Model Components

EQ2: Prevalence of Moldy Cannabis

Information estimating the probabilities of purchasing moldy cannabis products from tested and untested sources is captured in columns [B] and [C], and their difference is calculated in column [D]. 

The information collected suggests that most untested cannabis contains some amount of mold. We have no data on tested cannabis, but theoretically, most tested cannabis should be free of mold. There’s the complication that even cannabis that has been tested and shown to be mold-free may develop mold by the time the consumer ingests it. There’s also the complication that being contaminated isn’t an all-or-nothing phenomenon – we want to know not just whether or not the cannabis is contaminated but whether it’s contaminated enough to cause health effects when consumed.

Based on the context of the analysis being performed, the appropriate question to ask is: What’s the probability that the cannabis product at hand is contaminated enough that a consumer may get sick from consuming this product? Based on the information available, reasonable estimates for these probabilities are:

P(Mold if Untested) = 75% - 100%
P(Mold if Tested) = 0% - 10%

EQ3: Probability Consuming Moldy Cannabis Causes Illness

As with being contaminated, being ill is not an all-or-nothing phenomenon; rather, there are intensities of illness. As per the health effects presented in Figure 2, people may become mildly, moderately, severely, or direly ill, or they may die from ingesting moldy cannabis. We thus need to consider different intensities of illness in the calculations. These differences in the probabilities by severity of illness are captured in the model in columns [E] – [I]. Column [J] is the total probability that consuming a moldy cannabis product will make the user sick. 

At the same time, the probability that consuming a particular sample of moldy cannabis will cause illness is strongly dependent on the health status of the user. The model thus provides separate estimates in rows [1] – [3] based on the health status of the consumer, that is, respectively, for those who are healthy, who have a respiratory condition, and who are immunocompromised. 

EQ4: Costs of Illness

Figure 4 provides some estimates of costs associated with respiratory illnesses of varying severity. We also need to put a price on death (any value chosen is necessarily arbitrary). These costs are captured in row [4] of the model.

The total expected costs of mold-induced illness for people with each health status, column [K], aggregates the costs across the different severities of illness, using the respective probabilities as weights.

EQ1: Health Premium for Tested Cannabis

Finally, the value premium associated with consuming tested, rather than untested, cannabis products, column [L], is the product of the probability differential that the cannabis product is moldy, column [D], and the expected cost of mold-induced illness, column [K].

Model Insights

After playing around with different estimates for the inputs into the model (i.e., the blue boxes), I gained several insights.  

Relative Probabilities of Mold

First, what’s important is not the individual probabilities that tested vs. untested products contain harmful amounts, but rather the difference between the two. In other words, we get the same outcome if both tested and untested products tend to be moldy: 

[B] P(Mold if Untested) = 100% 
[C] P(Mold if Tested) = 50% 
⇒ [D] Difference = 50%

as we do if both products tend to not be moldy:

[B’] P(Mold if Untested) = 50% 
[C’] P(Mold if Tested) = 0% 
⇒ [D’] Difference = 50%

Consumers with Health Conditions

Second, generally speaking, with a reasonable difference in tested and untested probabilities of mold (column [D]),

  • Nonzero risks of death alone will generate a substantial (> $1,000) value premium (see row [3] in Figures 6) always enough to justify testing.
  • Nonzero risks of hospitalization alone will almost surely generate a value premium large enough to justify testing (see row [2] in Figures 6).

In other words, if users face any sort of dire health risk associated with consuming moldy cannabis, they should never use untested products. More generally, however, these users also face relatively high probabilities of becoming either mildly, moderately or severely ill after using moldy cannabis, which further justifies only using tested cannabis. In other words, the value of only using cannabis that has been tested for mold will always outweigh the expected costs for users who have respiratory conditions or who are immunocompromised (see Figure 5).

Figure 6A: High Testing Probability Differential

6a model sick 2 

Figure 6B: Low Testing Probability Differential

 6b model sick 2

Healthy Consumers

Healthy consumers who face only low risks of becoming mildly ill after ingesting moldy cannabis would probably have to have large costs of being mildly ill – say in terms of lost productivity – before the value to them of testing for mold would be worth the costs (see Figure 7). 

Figure 7: Healthy Consumers with Relatively High Time Costs

7 model healthy 2

What’s The Number?

I often conduct analyses for people who just want to know what The Number is. They don’t care about any of the details; they just want to know what the final number is. In this case, what they would ask for is, everything considered, what’s the overall value of testing for mold, the value that would appear in row [4], column [L] in my model?

I could calculate that number by figuring out the distribution of cannabis consumers by Health Status. The NIHsays that about 3% of the population is immunocompromised. VeryWellHealth says about 12% of the population has asthma, COPD, cystic fibrosis, or lung cancer. If we add another, say, 3% of the population to account for those with bronchitis and pneumonia, we get 15% of the population as having a respiratory condition. That leaves about 82% of the population as being healthy. If we then weight the value premiums (column [L]) by this population distribution (and assume that people who use cannabis are distributed similarly to the general population), we get an estimate of the overall value premium of about $1,500. 

HOWEVER, I’m reluctant to report this number, because I don’t think it’s particularly useful. People who want to know if buying tested cannabis is the right decision for them probably don’t 

care what the average cannabis user should decide about this issue; they care about whether it’s the right decision for them, given their particular situation, that is, given their health status. 

At the same time, when (only) The Number is reported, it often leads to the wrong decisions. In this case, someone who looks (only) at The Number would conclude that everyone should always pay the higher prices associated with testing for mold. However, the information I’ve collected suggests that most healthy people will probably suffer only minor effects, if any, from consuming moldy cannabis. In this case, healthy people may very rationally decide that the added costs are not worth the expected benefits.


I set out to estimate the value premium of consuming tested rather than untested cannabis in terms of the expected avoided costs from getting sick after consuming moldy cannabis. My analysis has led me to conclude:

  • Most untested cannabis is probably infected at least to some extent with mold, and some tested cannabis may also be infected.
  • The expected health effects of consuming a single mold-infected cannabis product:
    • May vary from mild to severe, and in rare cases may be fatal, and
    • Depend on health status, that is, whether the consumer is healthy, has a respiratory problem, or is immunocompromised
  • For the vast majority of cannabis users who are healthy, consuming mold-infected cannabis will probably only have minor effects, if any.
  • The expected health costs associated with consuming mold-infected cannabis will generally outweigh the added costs of buying tested cannabis products for consumers who have respiratory problems or who are immunocompromised. In other words, people with compromised health should generally spend the extra money and buy tested cannabis products.