Don’t Believe Everything You Read (On An I-502 Label): Washington Cannabis Potency Testing Found Unreliable

Can Washington consumers trust the label when they buy recreational cannabis? “No” is the answer, according to recent findings from in Scientific Reports, published in March.

The study, by Nick Jikomes and Michael Zoorob, entitled “The Cannabinoid Content of Legal Cannabis in Washington State Varies Systematically Across Testing Facilities and Popular Consumer Products” ( looked at the test results from the six busiest labs, according to test volume, in the state. These labs include Confidence Analytics (Lab A), Analytical 360 (Lab B), Green Grower Labs (Lab C), Integrity Labs (Lab D), Testing Technologies (Lab E), and Peak Analytics (Lab F).. When Nick Jikomes, of Leafly, and Michael Zoorob, of Harvard, analyzed the dataset of all flower and concentrates from the I-502 market, they discovered marked differences in the reporting of THC and CBD levels. The differences between reporting of the lowest and highest levels of THC showed from 17.7% to 23.2% median. Furthermore, the highest THC level was reported by Peak Analytics as 31.8% while the highest by Confidence Analytics was 27%.

Why is there such a discrepancy in the labs? There are two possibilities: either labs are getting the varying results due to systematic variances in material submitted or due to differences in testing procedure and standards.

Confidence Analytics Chief Science Officer Nick Mosley said that some variance could be equated to the use of either liquid or gas chromatography by a lab. In fact, Integrity Labs and Testing Technologies both use liquid systems and are the fourth and fifth highest scoring for potency in the study. The outlier in that scenario is Peak Analytics, however, and Mosley said that wouldn’t explain the variance in the other labs.

“There are big differences between labs in the cannabinoid measurements they report,” Jikomes said during a presentation on the data in March. “These differences cannot be explained by plausible confounds. This has more to do with the labs than what the labs are receiving.”

Before coming to that conclusion, Zoorob and Jikomes controlled the dataset to ensure that they were comparing testing from the same point in time, because the entire I-502 system saw increases in potency from 2014 to 2015. They also attempted to control for the variables of producer and cultivar/strain, while also noting that strain names are given by producers and don’t necessarily correlate to a real identification of a sample variety.

To fix the problem, Jikomes presented three metrics to test labs: a sensitivity test of ability to detect low levels of substances; consistency; or comparison of a variety of tests from the same lab, producer, and cultivar/strain and deviation, which could measure how far off from a system-wide baseline a lab is.

Preparing a sample of distillate for residual analysis.  ~ Confidence Analytics

Preparing a sample of distillate for residual analysis. 
~ Confidence Analytics

The dataset revealed that many of the labs didn’t reliably detect low levels of substances such as CBD. Either the samples of cannabis from these labs have often have 0% CBD or the testing technology isn’t accurate enough to identify lower rates, Jikomes explained. This would be a test of sensitivity of measurement.

A study by Vergara et al. in 2017 of cannabis potency in various states found that Seattle’s average potency was 19%, for example, and could be a good market for a reliable I-502 baseline.

Even beyond the information reported in this study, Confidence Analytics said they hope people understand that this is a complex issue and we need more, not less, testing to keep people protected. 

“It’s just a matter of safety to the consumer for the Washington State Liquor and Cannabis Board to mandate, or brands to offer, pesticide testing data,” Bobby Hines, Chief Technology Officer of Confidence Labs said. “We analyzed the costs per gram and it’s under a nickel for the complete battery of pesticide, terpenes, potency, water, solvents, and microbial.”

PanaceaAD quarter page - low.jpg

Last December a group of Canadian customers filed a class action lawsuit against Organigram, a Canadian producer,(  and now Xzibit the popular rapper and his Brass Knuckles cannabis oil line have been hit with a similar suit by California customers claiming pesticide poisoning. ( 

A nickel seems like a small cost to pay for public safety and peace of mind that you won’t be sued by a customer for tainted product.