Legal marijuana in the US may be less potent than packaging claims

A small sampling of marijuana sold legally in Colorado found most products contained substantially less tetrahydrocannabinol (THC) – the plant’s main psychoactive component – than was listed on the packaging.

“We know it’s happening and everyone’s talking about it, but no one has put the scientific rigour behind it,” says Anna Schwabe, who did the research while at the University of Northern Colorado but is now at cannabis company 420 Organics.

She and her colleagues obtained 23 samples from 10 Colorado dispensaries and had them tested for THC concentration at a private lab. Colorado, like other US states that have legalised recreational marijuana, requires dispensaries to label products with a potency range based on testing from third-party labs.

They found that samples were on average 23 per cent less potent than the low end of the range listed on the label. More than half of the samples were more than 30 per cent less potent.

“People are paying for a product and they’re not getting it,” says Schwabe. For medical marijuana users, this could mean someone not getting the proper dose, she says.

Discrepancies could come from labs sampling different parts of a plant, or using different testing equipment or methods, says Erik Paulson at Infinite Chemical Analysis Labs in California. But he says labs have a clear incentive to overstate potency. Buyers often pay more for marijuana with higher THC content, and dispensaries can pick the labs that give them the best numbers, he says.

In 2022, Paulson’s company analysed more than 150 samples from dispensaries in California, and found that nearly all had potencies that differed from the label by more than 10 per cent. He says reporting accurate, lower-potency results has caused his company to lose clients. “All of the producers know which labs to go with.”

Stephen Goldman at Kaycha Labs, a national testing company, says potency is especially prized in newer markets like California, Oregon and Michigan, and the results from Colorado don’t surprise him. “Could labs in general do better? Absolutely,” he says.

New Scientist contacted three of the dispensaries where samples appear to have been obtained based on licence numbers listed in the study. Two did not respond and one declined to comment.

Topics: