How strong is your weed, really? Scientists say labels often mislead

These findings come from a large-scale analysis of cannabis products purchased from licensed dispensaries across Colorado, the first state to legalize recreational marijuana. Published in Scientific Reports, the study represents the most detailed audit yet of legal-market cannabis labeling. It offers new insight into how potent today's cannabis has become, what testing and labeling practices are working, and where the system still falls short.

"Cannabis use has complex and wide-ranging effects, and we are working hard to better understand them," said senior author Cinnamon Bidwell, associate professor of psychology and neuroscience at CU Boulder and co-director of the Center for Health and Neuroscience, Genes and Environment (CU Change). "While that research plays out, we should, at the very least, be providing accurate information about the amount of THC in these products."

The project was funded by the Institute of Cannabis Research, Colorado's official state research institute, and carried out in partnership with MedPharm Research, LLC, a licensed testing facility, manufacturer, and retailer.

"We want to instill and foster trust in products, and the only way to do that is to continually evaluate and correct any issues that might be discovered," said Duncan Mackie, director of pharmacology at MedPharm and a co-author on the paper.

How the Research Was Conducted

Because federal law prevents university scientists from directly purchasing or handling cannabis from the legal market, industry collaboration was essential, explained first author Gregory Giordano, a professional research assistant in CU Boulder's Department of Psychology and Neuroscience.

A MedPharm "secret shopper" traveled throughout the state to collect 277 cannabis products from 52 dispensaries across 19 counties. The sample set included 178 flower products (loose flower and pre-rolled joints) and 99 smokable concentrates ranging from distilled liquids to Moroccan hash and waxy "sugar." Edibles were not part of this phase of research.

Photographs of each label were provided to Bidwell's team. The samples, identified only by number, were then tested by MedPharm chemists who did not see the original labels to ensure objectivity.

What the Tests Revealed

On average, cannabis flower products contained about 21% THC, or tetrahydrocannabinol, the main psychoactive compound in marijuana. Concentrates averaged 71% THC, with some reaching as high as 84%. By comparison, marijuana in the 1980s typically contained around 8% THC.

"THC content has increased significantly, and we know that greater THC exposure is likely associated with greater risks, including risk of cannabis use disorder and some mental health issues," notes Bidwell.

Products were considered "accurately labeled" if they contained within 15% of the THC amount shown on the label -- the same threshold the state uses.

About 44% percent of flower products failed to meet that standard, with 54 of those products inflating their THC content on the label and 23 containing more THC than the label indicated.

Either is concerning, the researchers say.

For those using cannabis medically, adequate dosing can be critical. For those using cannabis recreationally, taking more than expected can be dangerous.

Some discrepancies were large -- one flower product was labeled as having 24% THC but had only 16%. But on average, the difference between labeled and observed THC was about 2%.

Only four concentrate products were labeled inaccurately.

"When it comes to concentrates, I would say Colorado gets a good grade for labeling accuracy, but there are some real issues with flower," said Bidwell.

Why Potency Labels Can Differ

Previous research in other states has shown that third-party testing labs often inflate THC potency, possibly to gain the business of marketers wanting to attract consumers seeking stronger products.

However, there are other potential explanations for the discrepancy: Concentrates are often made from homogenous oils that are easier to analyze, whereas plants are inherently heterogenous and harder to test.

Small changes in testing protocols could reduce mislabeling, the researchers said.

Beyond THC: Other Cannabinoids Matter Too

The study also looked at several other cannabinoids (active compounds found in the Cannabis sativa plant), including cannabidiol (CBD), cannabigerol (CBG), and cannabigerolic acid (CBGA). Colorado law requires that companies put CBD levels on the label, a standard that 80% to 85% of products met.

But only 16% of products featured any information at all about other lesser-known cannabinoids.

Notably, CBG and CBGA, which have been associated with anti-inflammatory and anti-anxiety properties, was more abundant than CBD in products across categories.

"Focusing on THC on the label can actually do a disservice for consumers, because it creates an environment in which people buy based solely on THC content," said Bidwell. "Our data suggests that multiple other cannabinoids should also be reported on there."

Looking Ahead: Improving Cannabis Labeling

With continued support from the Institute of Cannabis Research, Bidwell and Mackie plan to expand their work to include edible cannabis products. As more states legalize marijuana for both medical and recreational purposes, they hope their research will help guide better regulatory standards.

"We all want the same thing," said Mackie, "a strong, successful industry that regulators can feel good about, businesses can thrive in, and customers can trust."