Hemp Seeds: THC Residue and Safety

Table of Contents

1. Where the THC Comes From
2. Reading the Numbers: Units and Scale
3. What Market Surveys Have Measured
4. EFSA's Exposure Assessment
5. Drug Tests: What Controlled Studies Found
6. Why a Screen Cannot Name the Source
7. CBD in Hemp Seed Foods
8. Hemp Seed Allergy and Cross-Reactivity
9. Oil Oxidation and Rancidity
10. Medication-Interaction Claims
11. Residues Passed Through Farm Animals
12. Who the Research Raises Concerns For
13. What Is Not Known
14. Connections
15. Key Research Papers
16. Featured Videos

Where the THC Comes From

The cannabis plant makes THC (delta-9-tetrahydrocannabinol, the compound behind the "high") and CBD (cannabidiol) inside tiny resin glands called trichomes. These glands are concentrated on the flowers and, to a lesser extent, the leaves. The U.S. Food and Drug Administration states that hemp seeds do not naturally contain THC or CBD, and that the trace amounts found in hemp seed ingredients come from contact with other parts of the plant during harvesting and processing (FDA).

In practice the combine harvester threshes flowers, bracts and seeds together. Sticky resin is smeared onto the outside of the shell, and dust from crushed flower material clings to it. A 2025 laboratory study measured the two parts of the seed separately and reported about 300 mg/kg of THC in the husk components against about 2 mg/kg in the kernel, a roughly 150-fold difference; the same study showed that THC could be selectively stripped from the outside of the seed with solvent and roasting steps without removing the husk (Kim 2025). That gap is why the form of the food matters:

• Whole, unhulled seed carries the shell and whatever resin is on it.
• Hulled seed ("hemp hearts") has had the shell mechanically removed, which takes most of the residue with it, though shell fragments and dust can remain if cleaning is incomplete.
• Hemp seed oil is pressed from whole seed. THC is fat-soluble, so resin on the shell can pass into the oil along with the seed's own fat. The surveys below consistently find higher concentrations in oils than in seeds.
• Protein powder is made from the pressed cake left after oil extraction, milled and sieved.

The plant matters too. "Hemp" is the name for low-THC cultivars of Cannabis sativa and "marijuana" for cultivars bred for high THC; the two belong to the same species but differ in chemical and genetic make-up and production practices (Salehi 2022). Low THC in the plant is not zero THC, which is why residue on the seed is possible at all.

Back to Table of Contents

Reading the Numbers: Units and Scale

Studies report THC in several units, and they are easier to compare once translated:

• 1 mg/kg = 1 µg/g = 1 part per million. The EU limit of 3.0 mg/kg for seeds means 3 micrograms of THC in each gram of seed.
• For oils, µg/mL and µg/g are close to equivalent, because a millilitre of oil weighs a little under a gram.
• A microgram (µg) is a thousandth of a milligram. EFSA's lowest effect level is stated in milligrams; residue studies work in micrograms.

The yardstick most often used for residue is EFSA's acute reference dose of 1 µg of THC per kilogram of body weight per day (EFSA 2015). An acute reference dose is the amount a regulator estimates can be taken in over one day without an appreciable health risk. EFSA derived it from the lowest dose found to produce effects in people, 2.5 mg of THC a day, divided by safety factors. For a 70 kg adult it works out to 70 µg in a day; for a 20 kg child, 20 µg.

Laid against that yardstick, using the highest values from the 2026 Turkish market survey described below (Kacargil 2026), simple arithmetic gives:

• 30 g of seed at the survey's highest seed value (2.08 µg/g): about 62 µg, close to the reference dose for a 70 kg adult and about three times it for a 20 kg child.
• 30 g of seed at the survey's mean seed value (0.78 µg/g): about 23 µg.
• One tablespoon (15 mL) of oil at the survey's highest oil value (31.08 µg/mL): about 466 µg, more than six times the reference dose for a 70 kg adult.
• One tablespoon of oil at the survey's mean oil value (6.64 µg/mL): about 100 µg.

These are illustrations of scale, not measurements of any product. They show why the oil, not the seed, is where most exceedances are found.

Back to Table of Contents

What Market Surveys Have Measured

The 1990s and early 2000s

The first wave of hemp foods in North America carried far more THC than today's. A U.S. Armed Forces Institute of Pathology laboratory summarised 79 hemp products it had tested, ranging from oils and nutrition bars to beer, tea and candy. THC ranged from none detected to 117.5 µg/g. The results were split into products tested before and after publication of a U.S. federal rule (21 CFR Part 1308) clarifying the listing of tetrahydrocannabinols: no THC was detected in 58% of the earlier group and in 86% of the later one, and the authors concluded that THC levels in marketed hemp products had fallen significantly compared with those available before 2003 (Holler 2008). An earlier paper from the same laboratory had already noted that THC in hemp oils had "been reduced considerably" since the first published studies (Bosy 2000). A 2001 study described 5 µg/g in oil and 2 µg/g in hulled seed as by then typical of Canadian hemp seed products (Leson 2001).

An Italian analysis of hemp foods from the same period found that THC content varied 50-fold between products, while CBD and the THC breakdown product CBN were absent from some samples and reached hundreds of nanograms per gram in others. Using the ratio of these compounds, the authors concluded that the majority of the specimens they analysed possibly originated from drug-type cannabis plants rather than fibre hemp (Pellegrini 2005).

The German national survey, 2008 to 2021

A German national survey analysed 511 samples across five groups of hemp foods: tea, seeds, seed oils, food supplements and non-alcoholic drinks. The authors estimated THC intake at average daily consumption and compared it with EFSA's acute reference dose and with the lowest dose reported to produce effects. Their finding: hemp beverages and seeds typically do not contain amounts of THC that can exceed toxicological thresholds. The problem products lay elsewhere. In hemp food supplements, including CBD products, 18% of samples had the potential to exceed the lowest effect level and 8% exceeded the minimum intoxicating dose, although THC in supplements fell significantly between 2018 and 2021. Tea made from hemp flowers was the other problem group (Dräger 2022).

The Turkish market survey, 2026

A forensic laboratory in Türkiye measured five cannabinoids in 12 hemp seed products and 20 hemp seed oils with a validated liquid chromatography–mass spectrometry method (Kacargil 2026):

• Seeds: delta-9-THC 0.11 to 2.08 µg/g (mean 0.78); CBD 0.13 to 2.40 µg/g (mean 0.88).
• Oils: delta-9-THC 0.11 to 31.08 µg/mL (mean 6.64); CBD 0.11 to 24.34 µg/mL (mean 5.15).
• Every seed and oil sample except one oil contained detectable total THC.
• The mean total THC across the oils, 10.21 µg/mL, was higher than the thresholds the authors list for Denmark (4 µg/mL) and the European Union (7.5 µg/g), and above the 10 µg/g or µg/mL they list for the USA, Canada, New Zealand and South Korea.

Every seed value in that survey sits below the EU seed limit of 3.0 mg/kg (Regulation (EU) 2022/1393); the oils are where the limits were crossed. The pattern across all these surveys is the same: wide variation between products, low values in most seeds, a tail of high values in oils, and the highest values of all in flower teas and cannabinoid supplements, which are not seed foods.

Back to Table of Contents

EFSA's Exposure Assessment

After its 2015 opinion set the acute reference dose, EFSA was asked to estimate how much THC Europeans might actually take in from hemp foods. Its 2020 report pooled 588 laboratory results for total delta-9-THC and used ordinary food categories (bread, pasta, oil and so on) as stand-ins for their hemp versions. It then modelled twelve separate single-food scenarios: what a person would take in on a day they ate that one hemp product (EFSA 2020).

The central finding was that the reference dose of 1 µg/kg body weight was exceeded in adult high consumers of most of the hemp products considered. The upper-bound estimates for adult high consumers, using the highest reliable percentile of THC measured in each product type, included (EFSA 2020):

• Hemp seeds: up to 9 µg/kg body weight (nine times the reference dose)
• Hemp oil: up to 21 µg/kg
• Bread and rolls made from hemp flour: up to 4.1 µg/kg
• Raw pasta: up to 6.4 µg/kg; fine bakery wares: up to 5.1 µg/kg
• Dietary supplements: up to 9.9 µg/kg
• Tea (infusion): up to 208 µg/kg
• Beer and beer-like drinks: up to 41 µg/kg
• Breakfast cereals, chocolate products, cereal bars and energy drinks: between 0.2 and 1.3 µg/kg

EFSA was explicit about the limits of these numbers. It named three main sources of uncertainty: the use of stand-in food categories for hemp products, the limited number of THC measurements, and analytical difficulties in quantifying THC. It concluded that the estimates "are expected to represent an overestimation of acute exposure" in the EU (EFSA 2020). In other words, the report describes the plausible worst day for a heavy eater of a high-residue product, not a typical day.

The European Commission subsequently set maximum levels for hemp seed foods, applying from 1 January 2023: 3.0 mg/kg for hemp seeds and derived products such as flour and protein, and 7.5 mg/kg for hemp seed oil (Regulation (EU) 2022/1393). Those limits apply to food sold in the EU. In the United States, the FDA's position rests on its 2018 evaluation of three GRAS notices for hulled seed, protein powder and oil, which it states contain only trace amounts of THC and CBD (FDA).

Back to Table of Contents

Drug Tests: What Controlled Studies Found

Urine drug testing for cannabis looks for THC-COOH, the inactive breakdown product the liver makes from THC. Testing usually happens in two steps: a cheap immunoassay screen, then a confirmation test by gas or liquid chromatography with mass spectrometry for any sample that screens positive. Each step has a cut-off concentration. The studies below report a 50 ng/mL screening cut-off as the one used in random drug screens (Bosy 2000) and 15 ng/mL as the confirmation cut-off used in U.S. federal drug-testing programs (Leson 2001; Steinagle 1999). Some studies also tested lower screening cut-offs of 20 or 25 ng/mL.

Read in date order, the controlled studies trace the fall in residue described above.

The 1997 studies: high-residue products

• Seed bars and cookies (Fortner 1997). Volunteers ate snack bars or cookies made from pressed hemp seed or hemp seed flour. After one bar, one specimen screened positive at a 20 ng/mL cut-off; after two bars, five did. Several specimens from people who ate three cookies screened positive at both the 20 and 50 ng/mL cut-offs, and some also tested positive on an FDA-approved on-site device. No specimen, however, produced a mass-spectrometry value that met reporting criteria; the highest measured was 3.1 ng/mL.
• Hemp oil (Costantino 1997). Seven adults each took 15 mL of a retail hemp oil. Of 18 urine samples collected afterwards, 14 screened above the 20 ng/mL cut-off, 7 above 50 ng/mL and 2 above 100 ng/mL, and THC-COOH was confirmed by mass spectrometry in every post-ingestion sample.

Around 2000: residue falling

• Hemp seed tea (Steinagle 1999). Twenty-two volunteers drank 12 or 24 ounces of hemp seed tea or a placebo tea. Ten specimens from seven people contained trace cannabinoids on mass spectrometry, all below the 15 ng/mL Department of Transportation cut-off, and none would have triggered a positive screen or confirmation.
• Retail hemp oils (Bosy 2000). THC-free volunteers took repeated 15 g doses of several commercial hemp oils. Peak THC-COOH in urine ranged from 1 to 49 ng/mL, and all volunteers were below the screening and confirmation cut-offs within 48 hours of stopping.
• Extended daily intake (Leson 2001). Fifteen THC-naïve adults took hemp oil blends every day over four successive 10-day periods, at daily THC doses from 0.09 to 0.6 mg. No one taking up to 0.45 mg a day screened positive at 50 ng/mL. At 0.6 mg a day, one specimen screened positive. The highest confirmed THC-COOH in any specimen was 5.2 ng/mL, below the 15 ng/mL federal confirmation cut-off. The authors calculated that 0.6 mg a day corresponds to about 125 mL of oil at 5 µg/g, or 300 g of hulled seed at 2 µg/g, and concluded that such levels appear low enough to prevent confirmed positives even with extended and extensive consumption.

Recent studies: low-residue products

• Low-THC oil and drug-driving tests (Hayley 2018). A double-blind crossover trial in 35 Australian adults, prompted by plans to allow low-THC hemp foods in Australia and New Zealand, compared a placebo oil with 5 mL doses carrying 0.046 or 0.092 mg of THC. No THC was detected in saliva, blood or urine at any time point. Trace THC-COOH appeared in blood four hours after the higher dose and in urine after both doses, far below any testing threshold.
• Twelve weeks of daily hemp seed (Baeck 2019). In Korea, volunteers ate 30 g of hemp seed a day (supplying about 53 or 125 µg of THC, depending on the product) or took two hemp seed oil capsules (about 11 µg), for one week or twelve weeks. All 64 short-term samples were below the 25 ng/mL screening cut-off. Of 480 long-term samples, the abstract reports 3 samples from 2 participants above the cut-off and also states that these samples were all negative in screening; and they concluded that products sold in Korea, eaten in normal amounts, did not produce positive urine tests.
• Oil, seed and skin application compared (Soyturk 2026). Fifteen volunteers in Türkiye took hemp oil by mouth, applied it to the skin, or ate hemp seed, with urine collected over 10 days. Oral oil produced the strongest signal: total THC-COOH was detectable in 72 of 77 samples (93.5%), with a highest median of 3.91 ng/mL and a maximum of 4.52 ng/mL. Eating seed produced recurrent but lower findings; skin application produced the weakest. The authors estimated a statistical upper limit of 9.35 ng/mL for the oral oil group and concluded that low-level urine results after hemp food call for cautious interpretation.

Across these studies, the outcome tracks the residue in the product. Products from the 1990s produced screen positives and, with oil, confirmed THC-COOH. Products at the residue levels described as typical after about 2000 produced detectable traces, sometimes in most samples, but at concentrations well below the confirmation cut-offs. None of this predicts the result for any particular product today, because residue still varies widely between batches, and oils at the top of the Turkish survey range carried far more THC than the oils in the recent trials.

Back to Table of Contents

Why a Screen Cannot Name the Source

THC-COOH is the same molecule whether the THC came from a joint, an edible or a hemp seed. A urine test measures how much is present, not where it came from, which is why testing programs rely on cut-offs high enough that trace exposures fall below them. The U.S. Department of Defense took a different route: in 2020 it barred service members from using hemp products, including hemp foods, stating that hemp products could cause a positive THC urinalysis and that product labels do not reliably state THC content (U.S. Navy ALNAV 069/20).

Forensic toxicologists are looking for ways to tell sources apart. The 2026 Turkish study found that after oral hemp oil, urine contained more of the delta-8 form of THC-COOH than the delta-9 form, at ratios from 1.92 to 8.83, and suggested that measuring the two forms separately may help interpret low-level results (Soyturk 2026). That is a research finding, not an established method used in routine workplace testing.

Poppy seeds pose the parallel problem for opiate tests: like hemp seeds, the seed itself makes none of the drug, the residue comes from the surface of the seed, and residue levels vary widely between batches. That story is told on the Poppy Seeds page.

Back to Table of Contents

CBD in Hemp Seed Foods

CBD products and hemp seed foods are different things. CBD oils and capsules are made from extracts of the flowers and leaves, where CBD is concentrated deliberately. Hemp seed foods are made from the seed, and any CBD they carry is the same kind of surface residue as their THC.

The amounts reflect that. In the Turkish survey, CBD in seeds ranged from 0.13 to 2.40 µg/g and in oils from 0.11 to 24.34 µg/mL (Kacargil 2026). An Italian analysis of 13 commercial hemp seed oils measured seven cannabinoids including their acid forms, of which CBDA (cannabidiolic acid) was the most abundant acid. Cannabinoid acids convert to the "active" neutral forms such as CBD with heat, light and air, and the authors studied that conversion to understand how storage conditions change the oil (Citti 2018). Older Italian work found CBD absent from some hemp foods and present at hundreds of nanograms per gram in others (Pellegrini 2005).

The German survey draws the contrast sharply: seeds and drinks typically stayed below toxicological thresholds for THC, while 18% of hemp food supplement samples, a group that includes CBD products, could exceed the lowest effect level for THC (Dräger 2022). On the regulatory side, the FDA's GRAS clearance covered hulled seed, protein powder and oil, not CBD; the FDA's 2023 statement on CBD is covered on the page FDA: CBD Needs a New Regulatory Pathway.

Back to Table of Contents

Hemp Seed Allergy and Cross-Reactivity

Hemp seed protein is often described as low in allergens, and one laboratory study supports a version of that claim: when hemp protein isolate was made from defatted seed cake and put through a simulated digestion, all known hemp allergens, including the main thaumatin-like protein and the lipid transfer protein, were eliminated by the isolation process, and no fragments of them survived digestion (Mamone 2019). That finding applies to a purified isolate tested in a laboratory model, not to whole seed eaten by people.

Clinical reports show that hemp seed allergy does occur:

• A 2003 report in the Journal of Allergy and Clinical Immunology describes anaphylaxis, a severe whole-body allergic reaction, after eating hempseed (Stadtmauer 2003).
• A 2019 clinical report describes cannabis and hemp seed allergy together (Alkhammash 2019).
• A 2022 case report describes allergic contact dermatitis, a skin reaction, from hemp seed oil applied to the skin (Clark 2022).
• A 2026 report describes allergy to the fatty part of hemp seed (Moncada Salinero 2026).

Which proteins are involved. A French study used blood from people sensitised to hemp seed and identified two seed storage proteins, vicilins and edestin (the seed's main protein), as potential hemp seed allergens. The same study found that sensitisation to hazelnut correlated with reactivity to a hemp seed extract rich in storage proteins, and that the hemp extract blocked 25 to 30% of hazelnut-sensitised people's IgE antibodies from binding hazelnut. The authors note that the clinical relevance of this possible cross-reactivity has yet to be tested in people with hazelnut allergy (Beriziky 2023).

Cross-reactivity with the cannabis plant and other foods. Cannabis plant allergy is a separate but overlapping problem. A lipid transfer protein named Can s 3 was identified in 2007 as a cannabis allergen (Gamboa 2007). Lipid transfer proteins are a family of plant proteins found across many plant foods, which is one route by which an allergy to one plant can extend to others. In a Spanish study of 545 people attending allergy clinics, 8.1% had a positive skin-prick test to cannabis leaf extract (14.6% among cannabis smokers), although symptoms directly attributed to cannabis were uncommon and mild; lipid transfer and thaumatin-like proteins were the allergens most often recognised (Larramendi 2013). A 2024 review lists five cannabis allergen components now catalogued by the international allergen nomenclature body and describes cross-reactivity syndromes mainly involving fruits, vegetables, nuts and cereals; it notes the usual presentation is hay-fever-type symptoms and skin swelling, but that cannabis allergy can also be life-threatening (Ebo 2024).

The research raises the question of cross-reaction mainly for people already sensitised to cannabis plant material, to lipid transfer proteins, or to tree nuts such as hazelnut. How often such people actually react to eating hemp seed has not been measured. More on the general topic is on the Food Allergy and Anaphylaxis pages.

Back to Table of Contents

Oil Oxidation and Rancidity

Hemp seed oil's fat profile is what makes it nutritionally interesting and also what makes it fragile. A Turkish analysis measured 54.85% linoleic acid (an omega-6 fat) and 18.13% alpha-linolenic acid (an omega-3 fat) in hemp seed oil, along with 707 mg/kg of gamma-tocopherol, a form of vitamin E that acts as a natural antioxidant (Özdemir 2021). Fats with several double bonds, like these two, react with oxygen easily.

A U.S. agricultural research laboratory study of hemp seed oil found that these polyunsaturated fats oxidise readily even at room temperature once in contact with air. Above roughly 100 to 140 °C, heat-driven breakdown of the fatty acids overtook ordinary oxidation. Lowering temperature slowed the reaction; adding 0.5% alpha-tocopherol lengthened the time before oxidation took off by up to 2.5 times; sealing oil droplets inside protein microcapsules, which kept oxygen away, lengthened it 7.3 to 9.3 times (Kim 2024). In the Turkish study, roasting seeds at 140 to 180 °C before pressing reduced the oil's tocopherols as roasting went on, while peroxide and p-anisidine values, two standard laboratory measures of oxidation, were generated from the early stage of roasting and varied significantly with roasting temperature and time (Özdemir 2021).

Oxidised oil develops the paint-like or fishy smell and bitter taste recognised as rancidity. No study of people eating rancid hemp seed oil turned up in the research for this article; the work above measures the oil, not health effects. The Flaxseed oil safety page covers the same chemistry in another omega-3-rich seed oil.

Back to Table of Contents

Medication-Interaction Claims

Hemp seed is sometimes said to interact with medicines. No clinical study of drug interactions with hemp seed food turned up in the research for this article. The interaction research that does exist concerns CBD and THC taken as products made to deliver those compounds, not the microgram-range residues measured in seeds and oils (see CBD in Hemp Seed Foods). Claims carried over from CBD products to hemp seed food are therefore not supported by evidence about the food itself. The absence of studies is not proof of safety either: it means the question has not been studied. Questions about any food or supplement alongside a prescription are for a clinician or pharmacist.

Back to Table of Contents

Residues Passed Through Farm Animals

Hemp seed cake, the residue left after pressing oil, is used as animal feed, and regulators have asked whether cannabinoids pass into meat and milk. In a U.S. study, 32 heifers were fed either a control diet or one that was 20% hempseed cake for 111 days. The cake averaged 1.3 mg/kg of combined CBD/THC. Neutral cannabinoids were not detected in the animals' blood or urine, but CBD/THC was found in fat at 6.3 to 10.1 ng/g at every withdrawal period tested, and the authors judged the chance of a person reaching the acute reference dose by eating that beef fat to be "remote" (Smith 2023).

Feed made from the whole hemp plant is a different matter. In a German experiment, dairy cows were fed industrial hemp silage (chopped whole plants, not seed cake); about 0.20% of the THC eaten passed into milk, and in exposure scenarios for milk and dairy products, the acute reference dose was exceeded in several consumer groups. The cows also showed changes in feed intake, milk yield, breathing and heart rates and behaviour (Wagner 2022). The contrast mirrors the human food story: the seed carries little, the flowering plant carries a great deal.

Back to Table of Contents

Who the Research Raises Concerns For

• People subject to drug testing. Controlled studies with today's lower-residue products found traces below confirmation cut-offs, but residue varies by batch, a test cannot identify the source, and the U.S. Department of Defense prohibits hemp foods outright for service members.
• Children. The acute reference dose is set per kilogram of body weight, so the same food is a proportionally larger dose for a small child. EFSA's assessment covered all age groups except infants (EFSA 2020).
• Pregnancy and breastfeeding. A 2022 review reports that THC crosses the placenta and has been detected in breast milk, and that cannabis use in pregnancy has been associated with adverse fetal outcomes and effects on child development (Lo 2022). That research concerns cannabis use, not hemp seed food; there is no specific safety research on hemp seed foods in pregnancy, which is why residue levels draw attention for this group.
• People with cannabis, lipid-transfer-protein or tree-nut allergies. Case reports describe anaphylaxis after eating hempseed, and laboratory work describes possible cross-reactivity with hazelnut (Stadtmauer 2003; Beriziky 2023).
• High consumers of hemp oil. In both EFSA's modelling and the market surveys, oil carried the highest residue of the seed foods (EFSA 2020; Kacargil 2026).

Decisions about diet in pregnancy, allergy and medication belong with a clinician; questions about drug-tested employment are governed by the employer's or agency's own policy.

Back to Table of Contents

What Is Not Known

• Current residue in the U.S. market. The most detailed U.S. product survey found in this research dates from 2008 (Holler 2008); no recent U.S. monitoring data on hemp seed foods turned up.
• Typical rather than worst-case intake. EFSA described its own estimates as overestimates built on stand-in food categories and limited data (EFSA 2020). Real-world eating patterns of hemp foods have not been surveyed directly.
• Long-term effects of repeated trace intake. The acute reference dose addresses a single day. No study has followed people who eat hemp seed foods daily for years.
• How common hemp seed allergy is. There are case reports and laboratory cross-reactivity studies, but no population estimate of how many people react to eating hemp seed.
• Drug tests with today's high-residue oils. The recent controlled trials used low-residue products. No controlled urine study has been published using oils at the upper end of the most recent survey ranges.
• Interactions. Drug interactions with hemp seed food, as opposed to CBD products, have not been studied.

Back to Table of Contents

Connections

• Seeds — the full seed collection
• Hemp Seeds — the main page, with legal limits and nutrition
• Hemp Seed Benefits — the deep-dive hub
• Protein and Digestibility — edestin, the seed's main protein
• Omega-3, Omega-6 and GLA — the fats that make the oil fragile
• Heart and Blood Pressure — the cardiovascular research
• Poppy Seeds — the parallel story of seed residue and opiate tests
• FDA: CBD Needs a New Regulatory Pathway — the regulation of CBD products
• FDA and FTC: Copycat Delta-8 THC Snacks — THC products that are not seed foods
• Food Allergy — how food allergy works
• Anaphylaxis — the severe reaction described after hempseed
• Contact Dermatitis — the skin reaction reported from hemp seed oil
• Hazelnuts — the nut with possible cross-reactivity
• Flaxseed: Oil, Cyanogens and Safety — oxidation in another omega-3 seed oil
• Urinalysis — what a urine test measures

Back to Table of Contents

Key Research Papers

1. Kim NP, Aditiya A, Min TJ. Unlocking safety: a mathematical model and selective extraction for Δ9-THC removal from hemp seed husks. J Cannabis Res. 2025;7(1):91. PubMed PMID: 41250237
2. Salehi A, Puchalski K, Shokoohinia Y, et al. Differentiating Cannabis Products: Drugs, Food, and Supplements. Front Pharmacol. 2022;13:906038. PubMed PMID: 35833025
3. Holler JM, Bosy TZ, Dunkley CS, et al. Delta9-tetrahydrocannabinol content of commercially available hemp products. J Anal Toxicol. 2008;32(6):428-32. PubMed PMID: 18652749
4. Pellegrini M, Marchei E, Pacifici R, et al. A rapid and simple procedure for the determination of cannabinoids in hemp food products by gas chromatography-mass spectrometry. J Pharm Biomed Anal. 2005;36(5):939-46. PubMed PMID: 15620517
5. Dräger H, Barthlott I, Golombek P, et al. Time Trends of Tetrahydrocannabinol (THC) in a 2008-2021 German National Survey of Hemp Food Products. Foods. 2022;11(3):486. PubMed PMID: 35159636
6. Kacargil CU, Soyturk H, Aydin AA, et al. Determination of cannabinoids in commercial hemp seeds and hemp seed oil products in Türkiye by LC-MS/MS. Forensic Sci Int. 2026;381:112836. PubMed PMID: 41621128
7. European Food Safety Authority (EFSA), Arcella D, Cascio C, Mackay K. Acute human exposure assessment to tetrahydrocannabinol (Δ9-THC). EFSA J. 2020;18(1):e05953. PubMed PMID: 32626501
8. Fortner N, Fogerson R, Lindman D, et al. Marijuana-positive urine test results from consumption of hemp seeds in food products. J Anal Toxicol. 1997;21(6):476-81. PubMed PMID: 9323528
9. Costantino A, Schwartz RH, Kaplan P. Hemp oil ingestion causes positive urine tests for delta 9-tetrahydrocannabinol carboxylic acid. J Anal Toxicol. 1997;21(6):482-5. PubMed PMID: 9323529
10. Steinagle GC, Upfal M. Concentration of marijuana metabolites in the urine after ingestion of hemp seed tea. J Occup Environ Med. 1999;41(6):510-3. PubMed PMID: 10390703
11. Bosy TZ, Cole KA. Consumption and quantitation of delta9-tetrahydrocannabinol in commercially available hemp seed oil products. J Anal Toxicol. 2000;24(7):562-6. PubMed PMID: 11043660
12. Leson G, Pless P, Grotenhermen F, et al. Evaluating the impact of hemp food consumption on workplace drug tests. J Anal Toxicol. 2001;25(8):691-8. PubMed PMID: 11765026
13. Hayley AC, Downey LA, Hansen G, et al. Detection of delta-9-tetrahydrocannabinol (THC) in oral fluid, blood and urine following oral consumption of low-content THC hemp oil. Forensic Sci Int. 2018;284:101-106. PubMed PMID: 29408718
14. Baeck S, Kim B, Cho B, et al. Analysis of cannabinoids in urine samples of short-term and long-term consumers of hemp seed products. Forensic Sci Int. 2019;305:109997. PubMed PMID: 31670217
15. Soyturk H, Goren IE, Atasoy Aydin A, et al. Low-level urinary cannabinoid findings after controlled administration of hemp seed products in Türkiye: Interpretive value of Δ8-/Δ9-THC-COOH patterns. Forensic Sci Int. 2026;387:113053. PubMed PMID: 42335722
16. Citti C, Pacchetti B, Vandelli MA, et al. Analysis of cannabinoids in commercial hemp seed oil and decarboxylation kinetics studies of cannabidiolic acid (CBDA). J Pharm Biomed Anal. 2018;149:532-540. PubMed PMID: 29182999
17. Mamone G, Picariello G, Ramondo A, et al. Production, digestibility and allergenicity of hemp (Cannabis sativa L.) protein isolates. Food Res Int. 2019;115:562-571. PubMed PMID: 30599980
18. Stadtmauer G, Beyer K, Bardina L, et al. Anaphylaxis to ingestion of hempseed (Cannabis sativa). J Allergy Clin Immunol. 2003;112(1):216-7. PubMed PMID: 12847507
19. Alkhammash S, Tsui H, Thomson DMP. Cannabis and hemp seed allergy. J Allergy Clin Immunol Pract. 2019;7(7):2429-2430.e1. PubMed PMID: 30910712
20. Clark E, Nilsson U, Samaran Q, et al. Allergic contact dermatitis from Cannabis sativa (hemp) seed oil. Contact Dermatitis. 2022;87(3):292-293. PubMed PMID: 35514096
21. Moncada Salinero A, Galán Díez T, Pineda de la Losa F, et al. Allergy to the Fatty Part of Hemp Seed. J Investig Allergol Clin Immunol. 2026; online ahead of print. PubMed PMID: 42825381
22. Beriziky P, Cherkaoui M, Linxe L, et al. Hemp seed: An allergen source with potential cross-reactivity to hazelnut. Food Res Int. 2023;169:112932. PubMed PMID: 37254358
23. Gamboa P, Sanchez-Monge R, Sanz ML, et al. Sensitization to Cannabis sativa caused by a novel allergenic lipid transfer protein, Can s 3. J Allergy Clin Immunol. 2007;120(6):1459-60. PubMed PMID: 17881043
24. Larramendi CH, López-Matas MÁ, Ferrer A, et al. Prevalence of sensitization to Cannabis sativa. Lipid-transfer and thaumatin-like proteins are relevant allergens. Int Arch Allergy Immunol. 2013;162(2):115-22. PubMed PMID: 23921252
25. Ebo DG, Toscano A, Rihs HP, et al. IgE-Mediated Cannabis Allergy and Cross-Reactivity Syndromes: A Roadmap for Correct Diagnosis and Management. Curr Allergy Asthma Rep. 2024;24(8):407-414. PubMed PMID: 38990404
26. Özdemir H, Bakkalbaşı E, Javidipour I. Effect of seed roasting on oxidative stability and antioxidant content of hemp seed oil. J Food Sci Technol. 2021;58(7):2606-2616. PubMed PMID: 34194096
27. Kim S, Hwang HS. Understanding the oxidation of hemp seed oil and improving its stability by encapsulation into protein microcapsules. J Food Sci. 2024;89(10):6321-6334. PubMed PMID: 39183680
28. Smith DJ, Serum EM, Winders TM, et al. Excretion and residue depletion of cannabinoids in beef cattle fed hempseed cake for 111 days. Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2023;40(4):552-565. PubMed PMID: 36897320
29. Wagner B, Gerletti P, Fürst P, et al. Transfer of cannabinoids into the milk of dairy cows fed with industrial hemp could lead to Δ9-THC exposure that exceeds acute reference dose. Nat Food. 2022;3(11):921-932. PubMed PMID: 37118216
30. Lo JO, Hedges JC, Girardi G. Impact of cannabinoids on pregnancy, reproductive health, and offspring outcomes. Am J Obstet Gynecol. 2022;227(4):571-581. PubMed PMID: 35662548

Regulatory sources: EFSA Panel on Contaminants in the Food Chain (CONTAM). Scientific Opinion on the risks for human health related to the presence of tetrahydrocannabinol (THC) in milk and other food of animal origin. EFSA Journal. 2015;13(6):4141. doi:10.2903/j.efsa.2015.4141. Commission Regulation (EU) 2022/1393 on maximum levels of delta-9-tetrahydrocannabinol in hemp seeds and products derived therefrom, EUR-Lex.

PubMed Topic Searches

1. PubMed: hemp foods and urine drug testing
2. PubMed: THC content of hemp seed oils
3. PubMed: hemp seed allergy and cross-reactivity
4. PubMed: hemp seed oil oxidative stability

Back to Table of Contents

Featured Videos