Many people within the cannabis industry would not have speculated earlier in their careers that they would one day find themselves working with a plant that was unlawful for most of their lives. But cannabis is special in that it wasn’t/isn’t just an illegal plant — cannabis has been proselytized against along the trajectory that has paralleled its prohibition.
With greater lucidity, many human beings have flung off the stench of bureaucratic hype that surrounded cannabis. Legality and the science that has been made possible have joined forces to squash decades of misinformation. The science requires people dedicated to accuracy, passionate about doing things right, and open minded enough to jump from any number of traditional ships to one traversing rocky waters with a crew of likeminded nonconformists. Luckily, our industry has attracted many of these types of scientists, like Cindy Orser, Ph.D.
I spoke with Dr. Orser, who serves as chief science officer at CLIP Labs, about her journey from more traditional industries to cannabis, as well as the current landscape of cannabis analytical testing.
Jason S. Lupoi: How did you end up working within the cannabis industry?
Cindy Orser: In 2014, a headhunter called me and asked if I would consider taking on the challenge of building out a cannabis testing lab for a public company in Las Vegas. At the time, I was winding down my own diagnostic company, ASDx Biosystems in Boulder, Colorado, where we developed rapid diagnostics for biological toxin subunits under a contract with the Department of Homeland Security in collaboration with the CDC [Center for Disease Control] for validation studies. The timing was perfect for me to take on a new challenge, which is what it turned out to be. At that time, few analytical instrument providers were keen on the fledgling cannabis testing industry and required some convincing to sell cannabis testing labs their instruments. In addition, there were no published methods on how to extract or analyze for cannabinoids or contaminants in cannabis matrices, and the instrument providers had not yet put any of their own technical talent to work on method development for the cannabis industry.
Once decisions had been made on which instrumentation provider would work with me, it was on to applying for the license, identifying auxiliary equipment, gas lines for endless compressed gas needs, air-handling needs, disposal of various waste materials, how to store retained cannabis material, security, and many more facility details. Once the instruments were qualified, a considerable amount of effort was expended on method development and locating certified reference materials and proficiency test providers for the various analyte categories. This presented its own challenges as the State of Nevada at that time would not allow labs to have any cannabis on site prior to obtaining their license. The surprise during the process of standing up a cannabis testing lab was my ability to attract exceptionally talented key scientific team members. Making strategic key hires proved critical to the build out of the laboratory, arduous method development, and the process of incorporating a Quality Management System and becoming ISO-17025 accredited. All in all, it turned out to be an incredible opportunity and accomplishment to add to my resume that has led to many follow-on possibilities, including doing it all again in another state for another entity.
JSL: As a prominent scientist in the industry, what are the top 3 things that you’ve learned about cannabis that you didn’t know prior to getting involved in the industry?
CO: To me, the most surprising aspect about cannabis that is still true today as it was 10 years ago is how little we understand about how the various secondary metabolites made by this amazing plant impart their physiological effects in a dose-dependent manner and on the flip side, how little is known about the detrimental impact of ready access to high-THC-content products and the lack of a conventional dose response in the case of CBD.
The second most astonishing fact about cannabis is just how many interesting and potentially beneficial secondary metabolites the plant can produce (estimated to be over 500) from thousands of years of anecdotal medicinal use. Chemistry ranges from the cannabinoid suite to caflanone, a flavonoid molecule with orphan drug status for pancreatic cancer, to cannabisativine, an alkaloid with analgesic activity found in the roots of cannabis.
Also surprising, the degree of genetic and phenotypic diversity that exists within the Cannabis sativa L. species, that can look like a corn stalk when grown for fiber and seed with very low cannabinoid content or resemble a Christmas tree in stature with upwards of 30% cannabinoid content when grown for resin and flower.
And sadly, that in California, up to 70% of the cannabis product sold emanates from the unregulated, untested, and robust black market. [Editor: a 2021 industry estimate, not a published figure.]
JSL: What do you think is the most cutting-edge analytical technology or technologies on the market right now for cannabis testing? Are there techniques you’ve seen used in other industries that might work for cannabis but haven’t yet caught on?
CO: I have recently become intrigued with the capabilities of Byers Scientific to monitor the complete gas-phase emission profile of the cannabis plant canopy to allow for more informed decisions about remediation to stay within air quality compliance standards. Through identifying and modeling canopy emissions in real time, Byers Scientific provides a factual, science-based assessment that benefits both cannabis cultivators and regulators.
I also don’t think we have seen the final application of mid-IR spectroscopy within the cannabis industry as a rapid analytical tool, such as the BSS series that Brian Smith, Ph.D., and Big Sur Scientific have patented that can distinguish drug-type THC cannabis from low-THC-content, resin- or fiber-type hemp.
There is an increasing interest in “authentication” of cannabis and cannabis products, and the industry should anticipate widespread non-targeted application of chemometrics in the years ahead with a focus on emerging real-world problems including quality control and fraud. A key to the success of its implementation will be the availability of data from samples used for modeling, external validation set data, as well as external raw data. To date, external raw data are rarely made available in other industries. Therefore, non-targeted screening technologies are currently access-limited due to the proprietary nature of samples stored within reference databases, e.g., within the wine industry. A more practical solution will be the creation of open-access databases as has been done for cannabis genomics data so that the full utility of non-targeted screening can be made widely available. A Canadian company called PURITY-IQ is doing just that, combining both genotyping data with NMR [nuclear magnetic resonance] fingerprinting.
JSL: What do you think is the biggest misconception about the cannabis analytics world?
CO: Unquestionably the biggest misconception about cannabis analytical testing is that a batch of flower has an absolute THC content and that no two labs agree on what that value is. A baby step toward eliminating this misconception would be implementing the requirement of reporting THC content as a range of values considering the measurement uncertainty (MU) budget as derived by each testing lab for every analyte and required by ISO-17025:2017. Uncertainty, in an analytical context, refers to the range around a reported result within which the true value can be expected at a certain probability. The heterogeneity of the cannabis plant contributes to the MU in the distribution of cannabinoids throughout the above-ground parts of the plant.
The US Department of Agriculture, in adopting their final hemp rules in the 2018 Farm Bill, got it right by requiring MU to be considered in determining if a sample qualifies as hemp, coming in under 0.3% THC. For example, if a laboratory reports a result as 0.35% THC with a measurement of uncertainty of +/− 0.06%, the distribution or range is 0.29% to 0.41%. Because the hemp definitional cutoff of <0.3% THC is within that distribution or range, the sample, and the lot it represents, passes as hemp. Both adult-use and medical cannabis regulated markets should acknowledge the MU ISO-17025:2017 requirement and let labs report analytes taking the MU range into account.
JSL: You’ve worked with a couple labs across different states. Where are the current regulations lacking and where have they succeeded?
CO: The cannabis industry would certainly benefit from a cohesive regulatory framework where all cannabis-complicit states agree to adopt the same contaminant monitoring list and action limits for pesticides, plant growth regulators, and heavy metals with penalties for violation. I have been fortunate to collaborate with a toxicology group at Arizona State University, led by Prof. Maxwell Leung and Prof. Tom Cahill, that will soon be publishing our study that reports on how disparate the contaminant monitoring regulations currently are and how the cannabis industry is contributing to our collective chronic exposure to environmental contaminants. [Editor: published in 2022; reference 8 and the addendum.]
Some labs adhere to the American Herbal Pharmacopeia for heavy metal tolerances, such as in Nevada, while most other cannabis-complicit states like California and Florida have adopted the US Pharmacopeia levels for inhaled heavy metals which are much more stringent.
We should be glad that the cannabis industry overall has created regulatory guidelines with a mandatory requirement for testing by independent third-party testing labs to provide for some level of self-monitoring in the absence of federal oversight.
JSL: You published papers regarding cultivar names versus chemistry, where differently named products had very similar chemistries. [1,2] Specifically in Nevada, 396 plant names boiled down to 3 chemistries. [2] Did those findings surprise you and your colleagues? Have you seen those same results across the industry?
CO: Our chemotypic data from Nevada demonstrated that cannabinoid data alone was not discriminatory of inferred groups, but instead, it was the terpenoid data, particularly the top four terpenes, beta-myrcene, beta-caryophyllene, limonene, and terpinolene, that were highly predictive of cluster membership in the large chemotypic dataset. We were not the first or last laboratory group to report the usefulness of terpenoid chemoprofiles in differentiation of cannabis cultivars when >95% are all drug-type THC cultivars; but we may have had the largest dataset at over 2,000 individual flower samples. Early pioneers in applying the utility of terpenoid chemometrics in clustering cannabis cultivars include Ethan Russo, MD, Justin Fischedick, Ph.D., and Mark Lewis, Ph.D.
The clustering of cannabis varieties using chemotypes is gaining in popularity. While we showed that one can reduce the number of terpenoids analyzed to just those four, the equipment, skill sets, calibration, and other technical requirements may hamper the application of chemotyping to cannabis clustering. In addition, the growth medium and environmental conditions are expected to influence quantity and quality of terpene expression, and thereby chemotypic data indirectly measures a genetically determined trait.
To bridge the chemoprofiling technical gap, we genotyped a subset of those 2,000 individual flower samples and identified 18 distinguishing single nucleotide polymorphisms (SNPs) out of the full genotyping dataset of 1,409 SNPs. [3] We proposed the use of this subset of 18 genetic markers in the form of optimized SNPs that can be typed in house with relatively inexpensive quantitative polymerase chain reaction systems. These assays can be used to predict the chemotype of a cannabis plant in its vegetative state without requiring a flower for chemotyping. An ancillary advantage of this approach is it provides an individual barcode in the form of a multi-locus genotype that can be used to identify and confirm the identify of a cannabis accession as a first step toward authentication.
Philip Henry and my group further reduced the set of 18 SNPs down to just 3 SNPs to create Tru-Hemp ID™, which can distinguish fiber hemp, resin-type hemp, and drug-type THC cannabis. [4] Since our reporting, terpenes have been in the spotlight for their utility in chemometrics as well as their not-so-subtle contribution to the overall distinctive physiological impact of cannabis varieties.
JSL: Quality is a simple word that carries very serious weight in cannabis testing. Where have you seen quality lacking within the industry from the lab vantage point? Where have some growers or product manufacturers gone wrong?
CO: There is a lack of quality in the cannabis industry from putting contaminant-laden product on dispensary shelves to wonton mis-labeling of products. Enforcement of existing state-by-state regulations is the only current option to improve quality without the industry stepping up to be ready when legalization happens. The other motivating pressure is class-action lawsuits; for example, California’s Prop 65 lists THC as a reproductive toxin with no safe harbor limits and actionable by the public, which should make everyone take notice especially when it comes to labeling products with “contains no THC,” something difficult to prove with no safe harbor limits.
The first federal oversight toward quality we are witnessing within the cannabis industry is coming through the hemp side with the USDA final rules published this year clinging to the <0.3% THC definition of hemp. [Editor: the federal definition moves to a total-THC standard on December 11, 2026; see the addendum.] The US Food and Drug Administration is considering cannabidiol (CBD) as a dietary supplement under new dietary ingredient (NDI) requirements in the face of CBD’s prescription drug status under Epidiolex® with little concrete discussion of CBD in foodstuffs at the federal level. [Editor: in January 2023 the FDA concluded the supplement framework is “not appropriate” for CBD; see the addendum.]
CLIP Labs along with two other California-based cannabis testing laboratories have recently formed a non-profit called QUALITYSAFE™ to serve as a beacon of optimism that the industry can move toward a more consumer focused outlook given the inevitable move toward federal legalization and establishment of cohesive, uniform food and drug safety requirements. The dietary supplement and food manufacturing industries have a lot to offer the cannabis industry through adopting hazard analysis critical control points (HACCP). The principles are well established: performing hazard analyses, identifying and monitoring critical control points (CCP), establishing corrective actions, verification procedures, and record keeping. Being ready for federal legalization through adopting HACCP would reduce product loss, increase product consistency and quality, provide inventory control, and increase profits.
JSL: Looking back on your career and all your experiences, what do you think the cannabis industry offers that may not have been previously attainable to you?
CO: Entering the cannabis industry early in the evolution of science-based independent laboratory testing has given me the opportunity to inform policy, write protocols, and have a voice and platform to share my vision and promote my ideas. The combination of my curiosity and previous broad scientific experience with being at the forefront of cannabis analytical testing has afforded me a choice of projects to pursue from day one, from beta-testing new technology, formulation work, applying chemometrics to the cultivar-naming game, and ending up with three patent applications from terpene applications to SNPs to date.
Editor’s Addendum (October 2026): What Has Changed Since This Conversation?
Measurement uncertainty is now federal policy. Orser’s example in answer four, a lot reading 0.35 percent THC with a ±0.06 percent uncertainty counting as hemp because its range includes 0.3 percent, is how the USDA’s hemp program has worked since the final rule took effect in 2021: a lot complies if the laboratory’s reported total THC, with its measurement of uncertainty, gives a range that includes 0.3 percent, and the rule’s definitions use her exact numbers as the worked example. The negligence threshold was raised from 0.5 to 1.0 percent, and sampling moved to a 30-day pre-harvest window.[5] The next change is the one that matters for products rather than crops: from December 11, 2026, the federal hemp definition shifts to a total-THC standard that includes THCA, with a cap of 0.4 mg of THC per container for finished goods.[6] Our hemp redefinition explainer covers it. Adult-use and medical states still mostly print a single THC number; the range reporting she asked for has not arrived there.
Rapid spectroscopic potency is closer, and still a screen. Orser flagged mid-infrared instruments that sort drug-type from fiber-type cannabis. The most rigorous recent test of the approach is a 2025 near-infrared study of 75 flower samples that classified high- versus low-THC material with 96.7 percent accuracy on a held-out test set and predicted THC in low-range samples to within about 0.07 percentage points. Its cutoff was 2 percent THC, not the 0.3 percent hemp line, two CBD-rich samples confused the model, and the authors describe the method as a first-line screen that still needs confirmatory chromatography.[7] That is also where regulators are: compliance testing remains chromatographic, at accredited laboratories. Lupoi’s own look at NIR chemometrics, Spectroscopic Analysis of Hemp Flour, makes the same point Orser does about shared validation data.
The contaminant study was published, and it was worse than she implied. The Arizona State collaboration Orser mentions in answer five appeared in Environmental Health Perspectives in 2022, with Orser among the authors. Across 36 states and the District of Columbia, regulators listed 679 different contaminants, 551 of them pesticides, and the action limits for the same contaminant varied by up to four orders of magnitude between jurisdictions. The authors called for national guidelines built on conventional risk assessment.[8] None has been adopted. Our contaminant testing pillar tracks the state-by-state picture.
The SNP work became a paper. The genotyping she describes in answer six was published in 2020: 23 single-nucleotide markers run on 420 accessions from Canadian and Nevada cultivators, separating fiber hemp from resin-type cannabis and resolving four resin-type clusters that line up with terpene chemotypes.[4] The Nevada chemotype papers it grew from are references 1 and 2 below; our article on why cultivar names are mud tells the wider story, and why chemotyping matters covers the method.
And the FDA answered the CBD question. In answer seven Orser noted the FDA was weighing CBD as a dietary supplement. In January 2023 it concluded that the existing frameworks for foods and supplements are “not appropriate” for CBD and that a new pathway would require Congress.[9] Our CBD profile has the rest.
References
- Orser C, Johnson S, Speck M, Hilyard A, Afia I. Terpenoid chemoprofiles distinguish drug-type Cannabis sativa L. cultivars in Nevada. Natural Products Chemistry & Research. 2018;6:1–7.
- Reimann-Philipp U, Speck M, Orser C, et al. Cannabis chemovar nomenclature misrepresents chemical and genetic diversity; survey of variations in chemical profiles and genetic markers in Nevada medical cannabis samples. Cannabis and Cannabinoid Research. 2020;5(3):215–230.
- Henry P, Hilyard A, Johnson S, Orser C. Predicting chemovar cluster and variety verification in vegetative cannabis accessions using targeted single nucleotide polymorphisms. PeerJ Preprints. 2018.
- Henry P, Khatodia S, Kapoor K, et al. A single nucleotide polymorphism assay sheds light on the extent and distribution of genetic diversity, population structure and functional basis of key traits in cultivated North American cannabis. Journal of Cannabis Research. 2020;2(1):26.
- U.S. Department of Agriculture. 7 CFR § 990.1, Definitions: acceptable hemp THC level; measurement of uncertainty; total THC. Establishment of a Domestic Hemp Production Program, final rule, Federal Register document 2021-00967, January 19, 2021.
- Congressional Research Service. Changes to the statutory definition of hemp and implications for agricultural policy. In Focus IF13136, updated September 28, 2026.
- Mistek-Morabito E, Wilson WB, Lednev IK, Urbas AA. Analysis of cannabis plant materials by near-infrared (NIR) spectroscopy and multivariate data analysis for differentiating low-THC and high-THC cannabis. Forensic Chemistry. 2025;46:100698.
- Jameson LE, Conrow KD, Pinkhasova DV, et al. Comparison of state-level regulations for cannabis contaminants and implications for public health. Environmental Health Perspectives. 2022;130(9):097001.
- U.S. Food and Drug Administration. FDA concludes that existing regulatory frameworks for foods and supplements are not appropriate for cannabidiol, will work with Congress on a new way forward. Statement, January 26, 2023.
Originally published May 18, 2021. Editor’s note, addendum and reference updates added October 2026. Dr. Orser’s answers are reproduced as given; bracketed notes are the editors’.
