Marijuana-infused cookies sit on the packaging table at edibles manufacturer The Growing Kitchen in Boulder on Sept. 26, 2014. (Brennan Linsley, The Associated Press)

Lab: Edibles’ potency more accurate than a year ago, but buyer beware

Cannabis consumers in Colorado can trust the potency claims printed on the packaging of marijuana-infused edibles more than they could one year ago, according to new lab testing data commissioned by The Denver Post.

In comparing the claimed amounts of psychoactive component THC versus actual amounts in 10 different edibles, The Post learned that infused products have become more reliable than they were in March 2014, when the paper conducted a similar series of lab tests.

But there’s still plenty of room for improvement, says the scientist who conducted the tests.

“These results are better than last year’s test because there is nothing here that has no THC and nothing being sold that has over-the-limit THC,” said Remy Kachadourian, the lab director at Steep Hill Halent of Colorado, which conducted the tests for The Post via two area pot shops’ licenses. “That’s big progress, but it’s not totally there yet.”

SPECIAL REPORT: How do you stop a consumer from eating too much? More than 15 months into recreational sales, edibles safety still concern for Colorado

In Colorado, edibles are packaged primarily by THC amounts. A single-serving cookie will be labeled for 8 to 10 milligrams of THC; a package of gummies or a chocolate bar bought recreationally often will be labeled for 80 to 100 milligrams — or eight to 10 individual servings.

When The Post conducted a similar test on edibles in March 2014, some of the results embarrassed the then-new industry. Four different products from edibles company Dr. J’s, each labeled for 100 milligrams of THC, tested between 0.2 and 5 milligrams of total THC. Popular chocolate company Incredibles tested over in the 2014 test, 146 milligrams of THC on a product labeled for 100.

Both Dr. J’s and Incredibles fared better in the 2015 tests, although their progress still shows room for improvement. A Dr. J’s dark chocolate StarBarz labeled for 74.3 milligrams of THC tested for 52.2, or 29.7 percent lower. An Incredibles Monkey Bar labeled for 100 milligrams of THC tested for 82.9, a 17.1 percent difference.

“A 10 percent variability is pretty much exact for me,” said Kachadourian, noting that Colorado is looking to create statewide laboratory standards for its 18 state-licensed marijuana testing labs. “Something testing for 20 percent under, that’s OK. When it’s more than that, it can be a big problem.”

GRAPHIC: View the 2015 analysis of 10 popular Colorado edibles

Other lab testing experts agreed with Kachadourian about the inexact nature of testing pot, especially given the varying methodologies and lab-to-lab results.

“It’s not an exact science,” said Genifer Murray, founder and president of CannLabs.

While edibles testing isn’t exact, neither is the dosing of edibles, according to Dr. Alan Shackelford, a Denver-based medical marijuana doctor who is regularly conducting cannabis studies in Israel: “Even though the extraction methods and production methods are very refined and have seen significant improvement since the pot brownie of the ’60s, (edibles are) still a nonstandardized way of administering cannabis.”

The new lab test’s underperformers, Kachadourian said, were Dixie Elixirs and EdiPure, which each tested at around 55 percent less THC than what their products were labeled for.

Kachadourian said some products, including the infused drinks Dixie sells, can be more difficult to test than others.

“When you take oily cannabinoids and put them in water, it’s like putting olive oil in water,” Kachadourian said. “It’s hard to get it to be homogenous. I don’t know their process, but that might explain the lower number. But we did repeat the test, and the number came out almost exactly the same.”

When contacted for a comment on their product’s low test results, Dixie conducted its own test at a different state-licensed lab on what they say was the same batch and product from the same shop to results that were much closer to their labeled amount. The varying results, the company says, speak to the state’s need for a reference lab.

“We literally went down to Denver Relief and found the same flavor from the same batch,” said Dixie chief marketing officer Joe Hodas, who forwarded a new test from CMT Laboratories showing a result of 80.9 milligrams of THC on a product labeled for 75. “There’s not a primary reference lab that an (independent) lab can go to and see, ‘How do we extract THC with a product that includes these ingredients?’ They have to make up their own methodologies based on their own knowledge.

“It’s less an issue of our kitchen and more an issue of the labs.”

Ricardo Baca: 303-954-1394, or

This story was first published on