How to Measure THC Content and Potency

THC content in cannabis is measured primarily through laboratory chromatography, with high-performance liquid chromatography (HPLC) considered the gold standard because it can detect both THC and its raw acidic precursor, THCA, without altering them during analysis. But “measuring potency” is not as clean-cut as it sounds. The number on a dispensary label, the result from a field test kit, and what your body actually experiences can be three very different things, and the gaps between them are wider than most consumers realize.

The Two Main Lab Methods

Professional cannabis testing labs rely on two core techniques: high-performance liquid chromatography (HPLC) and gas chromatography coupled with mass spectrometry (GC-MS). Both separate a cannabis extract into its individual chemical components so each one can be identified and quantified, but they handle the plant’s chemistry differently, and that difference matters for accuracy.

HPLC works at room temperature. A liquid solvent carries the dissolved cannabis extract through a packed column, and a detector reads each compound as it emerges. Because nothing gets heated, HPLC preserves the acidic forms of cannabinoids, particularly THCA, which is the form actually present in raw plant material. That is important because THCA converts to THC when heated, so knowing how much of each exists in the flower tells you both what is physically there and what would become active if you smoked or vaped it. HPLC methods have been validated for simultaneously quantifying THC, THCA, CBD, CBDA, CBG, CBGA, and CBN in plant material.¹

GC-MS, by contrast, vaporizes the sample using high heat before it enters the column. That heat converts THCA into THC during the analysis itself, so the instrument reports a single combined THC number. In theory this makes things simpler, but in practice it introduces complications. A 2025 study found that a plant matrix effect significantly inflated THCA and THC readings in cannabis extracts compared to calibration solutions, because the compounds interact with active sites on heated inlet surfaces. The researchers proposed using “analyte protectants” to block those interactions and bring the numbers back in line.² This kind of systematic overcount is one reason many regulatory frameworks have moved toward HPLC-based methods.

For consumer products like oils, gummies, and tinctures, labs often use liquid chromatography paired with tandem mass spectrometry (LC-MS/MS), which combines the gentle separation of HPLC with extremely sensitive detection. One validated LC-MS/MS method achieved a lower limit of quantification as low as 0.195 ng/mL for THC and CBD in oil matrices, meaning it can accurately measure cannabinoids even at trace concentrations.³

Why Sample Preparation Can Make or Break the Result

Before any instrument can read the cannabinoid content, the THC has to be pulled out of the plant material and dissolved into a form the machine can work with. This extraction step sounds routine, but the choice of solvent and technique meaningfully changes what gets measured.

Different solvents show different selectivity for different cannabinoids. Methanol tends to be efficient for extracting a broad range of cannabinoids, while ethyl acetate and certain hexane blends outperform pure hexane specifically for THC-type compounds. Ethanol, despite being widely used, has been found generally inferior for cannabinoid extraction compared to these alternatives.⁴ This means two labs using different solvents on the same batch of flower can arrive at different potency numbers, even if their instruments are perfectly calibrated.

Extraction techniques also vary. Ultrasonic-assisted extraction uses sound waves to create microscopic bubbles in the solvent, which collapse and drive the liquid deeper into plant cells to pull out more cannabinoids. Factors like particle size, temperature, and how long the sample is exposed to the sound waves all affect the final yield.⁵ Other labs use simpler dilution-and-shaking approaches. The point is that “THC content” is not purely a property of the flower; it is partly a property of the method used to measure it.

Where You Sample the Plant Changes the Number

Cannabis plants are not uniform. THC concentration varies from the top of the plant to the bottom, between individual plants of the same cultivar, and even within a single bud depending on which tissue you test. A study of greenhouse-cultivated cannabis found that total THC varied by about 3 to 7 percent of actual content within a single stratum of the plant across multiple batches, and between the top and bottom of plants, THC varied by roughly 5 to 6 percent. Even between plants of the same cultivar grown together, average THC differed by about 2.8 percent, a statistically significant gap.⁶

The tissue type matters as well. Bracts, the small leaf-like structures surrounding the seed cavity, tend to have the most consistent trichome distribution and the highest cannabinoid concentration, making them the most reliable sampling unit. Sugar leaves and calyxes are also analyzed, but their shapes vary enough that results need more cautious interpretation.⁷ If a lab pulls its sample from a dense, trichome-heavy top cola and a regulator pulls from a lower branch, the numbers will differ even though both are technically correct for the material they tested.

The Label Accuracy Problem

For consumers buying legal cannabis, the potency number printed on the package is often the only data point available. It is also frequently wrong, and not by a small margin. A study that independently tested retail cannabis flower found the average observed THC potency was about 23 percent lower than the lowest values reported on labels and nearly 36 percent lower than the highest label values. Roughly 70 percent of samples tested more than 15 percent below the label claim, and three samples contained only half the maximum THC potency listed on their packaging.⁸

This is not just random error. An analysis of Washington State testing data found that median total THC for the same type of flower product ranged from about 17.7 percent to 23.2 percent depending on which lab did the testing. Every pairwise comparison between labs was statistically significant. The researchers identified both cannabinoid inflation and varying sensitivity to low-level cannabinoids as contributors to the systematic differences.⁹

Several forces drive this inflation. Producers have a financial incentive to market high-THC numbers because consumers tend to equate potency with quality and are willing to pay more. “Lab shopping,” where a producer sends samples to whichever lab returns the highest numbers, has been widely reported across legal markets. Underneath all of this is a lack of standardized testing protocols. Different labs use different extraction solvents, different instruments, different calibration materials, and different sampling procedures, and these procedural differences compound.

Calibration and Reference Material Gaps

One underappreciated reason for lab-to-lab disagreement is the quality of certified reference materials (CRMs) that labs use to calibrate their instruments. A reference material is a chemical standard of known concentration against which the lab compares its sample readings. If the reference material itself is slightly off, every measurement drifts. Research into this issue has identified certified reference materials as the single largest contributor to measurement uncertainty in cannabis testing, and it is a structural problem tied to flaws in how those standards are produced and certified.¹⁰

Moisture content adds another layer of confusion. Legal thresholds for hemp, which in the United States is defined as cannabis containing no more than 0.3 percent total THC on a dry-weight basis, depend on properly accounting for how much water is in the sample. Normalizing THC content to dry weight makes sense when comparing materials with very different moisture levels. But a study found that when hemp is dried to under about 13 percent moisture, the advantage of dry-weight normalization largely disappears because the variability between labs in reporting THC on a dry-weight basis is already so high.¹¹ For hemp farmers, this means a crop hovering near the 0.3 percent legal limit can test legal at one lab and illegal at another, with real financial consequences.

Rapid and Field-Based Testing

Not every situation calls for a full lab analysis. Law enforcement, agricultural inspectors, and growers doing quality checks often need a quick answer in the field. Several portable technologies have emerged to fill this gap, though they trade precision for speed.

Handheld near-infrared (NIR) spectrometers are among the most promising. These devices shine infrared light through a cannabis sample and analyze the reflected spectrum, which correlates to the sample’s chemical composition. A battery-operated Fourier-transform NIR device was shown to quantify major cannabinoids in hemp in roughly 15 seconds, with prediction models showing strong correlations to lab reference values and a low standard error of prediction.¹² A cloud-connected handheld NIR device has also demonstrated the ability to distinguish CBD-type from THC-type cannabis in the field, with spectra that closely matched results from laboratory HPLC.¹³

At the even simpler end, thin-layer chromatography (TLC) combined with smartphone-based color analysis offers a low-cost semiquantitative screening method. A modified TLC plate with a silver-ion retention zone can separate THC-type compounds from CBD-type compounds within minutes, and a smartphone camera can estimate the THC percentage from the resulting color spots, with detection limits in the nanogram range.¹⁴ These methods are not precise enough for regulatory compliance testing, but they can answer the practical question of whether a sample is high-THC or low-THC cannabis on the spot.

Measuring THC in Edibles and Complex Products

Flower is relatively straightforward to analyze compared to edibles. When THC is dissolved into butter, infused into chocolate, or suspended in a gummy, the surrounding food matrix interferes with extraction and detection. Fats, sugars, emulsifiers, and dyes all complicate the chemistry, and different product types require different cleanup steps before the cannabinoids can be isolated and measured.

A comprehensive review of analytical methods for cannabis edibles found that a wide range of extraction and cleanup techniques are used depending on the matrix. Oils require different handling than candies, beverages differ from baked goods, and even pet food products present unique challenges. The analytical techniques themselves span LC, GC, direct analysis in real time (DART), and various forms of mass spectrometry.¹⁵ The practical takeaway is that potency testing for edibles is harder, more variable, and more prone to error than for flower, which is why homogeneity testing (checking that THC is evenly distributed throughout a batch of gummies, for instance) is a regulatory requirement in many jurisdictions.

Telling THC Isomers Apart

The rise of delta-8-THC, delta-10-THC, and other structural isomers has created a new measurement headache. These molecules are chemically similar to delta-9-THC, the primary psychoactive form, but differ slightly in the position of a double bond. Because they are so structurally close, they can overlap or interfere with each other during chromatographic analysis. Research into cannabinoid detection in hair specimens found that exo-THC elutes just before delta-9-THC, and at very high concentrations, the two may not separate cleanly, preventing accurate measurement of either.¹⁶

For consumers, this means a product labeled as “delta-8 only” might contain trace delta-9-THC that went undetected because the lab’s method could not cleanly separate the two. For forensic and regulatory labs, distinguishing between isomers is critical because legal status often hinges on which specific THC variant is present. Labs addressing this challenge tend to use longer chromatographic columns, slower flow rates, or MS/MS detection to achieve adequate separation.

How THC Degrades Over Time

THC is not stable. It breaks down over time, and the rate depends on temperature, light, and pH. The primary degradation pathway converts THC into cannabinol (CBN), a much less psychoactive compound. A four-year study tracking hashish and marijuana stored under different conditions found that nearly 100 percent of THC degraded over that period. The conversion followed first-order kinetics, meaning the breakdown accelerates early and slows as less THC remains. Temperature primarily affected the speed of conversion, while light exposure changed both the speed and the ratio at which THC converted to CBN.¹⁷

In liquid preparations like cannabis resin solutions, the transformation rate climbs substantially as temperature rises above about 70°C, and acidic conditions (low pH) accelerate the breakdown further. THC is most stable between roughly pH 4 and 12, meaning strongly acidic environments are particularly destructive.¹⁸ This matters for anyone storing cannabis products long-term, and it also explains why old test results may not reflect what is currently in a product. A flower tested at 25 percent THC that sits on a shelf for a year in warm, light-exposed conditions will have meaningfully less THC than the label claims.

Measuring THC in the Human Body

Potency testing on a product is one thing; measuring how much THC is in your body is a different analytical problem. Biological testing for THC is done on blood, urine, oral fluid, and hair, and each matrix tells a different story.

Oral fluid has gained traction as a preferred sample type for detecting recent cannabis use because THC appears in saliva quickly after smoking or ingestion and clears relatively fast, making it a better indicator of impairment timing than urine, which reveals only a history of past consumption. Specific collection devices have been developed to standardize oral fluid sampling, and research interest in point-of-care oral fluid biosensors has grown substantially over the past two decades.¹⁹

For forensic and clinical purposes, validated LC-MS methods can simultaneously quantify THC and its metabolites, including THC-COOH (the breakdown product tested for in urine drug screens) and 11-OH-THC (the active metabolite produced in the liver), across oral fluid, urine, and whole blood from a single analytical platform.²⁰ The challenge here is not just detecting THC but interpreting what the level means. Blood THC concentration drops rapidly after inhalation, and factors like body fat percentage, frequency of use, and individual metabolism all affect how long the compound lingers.

Why the Percentage on the Label Does Not Predict Your Experience

Even a perfectly accurate THC percentage is a limited predictor of how a cannabis product will affect you. The same dose of oral THC produces different blood levels depending on whether you have eaten recently. A study of oral THC capsules found that consuming them with high-fat food significantly increased both the peak blood concentration and the total exposure over 24 hours compared to taking them on an empty stomach. The study also found that females achieved a significantly higher peak blood concentration than males after the same 5 mg dose in a fasted state.²¹

Beyond individual biology, there is growing recognition that THC percentage alone does not capture how a given cannabis product will perform therapeutically or recreationally. A chemovar indexing system that incorporated both primary cannabinoid percentages and terpene content identified 478 distinct chemical profiles among cannabis products. Different chemovars varied in treatment effectiveness and reported side effects, suggesting the terpene composition alongside THC and CBD levels contributes to the overall experience.²² In other words, two flowers that test at the same THC percentage but have different terpene profiles can feel quite different to the user. The industry’s fixation on THC percentage as the main quality metric is understandable given that it is the easiest number to put on a label, but it tells an incomplete story. The analytical tools exist to characterize cannabis products much more fully; the market just has not caught up yet.