When a hemp seed is pressed for oil, roughly two-thirds of what went into the press comes back out as a dry, protein-dense cake. For decades that cake went to compost, cheap fertiliser or the back of a barn. Now it is being tested as a serious replacement for soybean meal, cottonseed meal and fish meal, and the results keep pointing the same way: the animals eat it, grow on it and show no sign of harm. Earlier this year the evidence came from a fish tank. Now it comes from a sheep pen.
The new data come from a study published on 12 August 2026 in BMC Veterinary Research by Başak Özten, Hacer Tüfekci and Hüseyin Çayan of Yozgat Bozok University and Kırşehir Ahi Evran University in Türkiye. The team fed 40 Akkaraman lambs rations containing 0%, 5%, 10% or 15% hemp seed meal for 75 days and tracked live weight, feed intake, body measurements and blood chemistry throughout. The full paper is open access at https://doi.org/10.1186/s12917-026-05807-y. Below we walk through what it found, place it next to the European perch trial this article originally covered, and explain why the regulatory answer to “can farmers actually use this?” still differs sharply between Europe and the United States.
Key Takeaways
Forty male Akkaraman lambs, starting at an average 27.63 kg, were fattened for 75 days on rations in which hemp seed meal replaced part of the sunflower and cottonseed meal at 5%, 10% or 15% of dry matter. Final live weights were 46.17 kg on the control ration and 46.37, 47.23 and 47.43 kg on the three hemp rations, a difference that was not statistically significant. Total weight gain, daily gain over the full period, feed conversion ratio and every body measurement were likewise indistinguishable between groups. Blood chemistry stayed within normal ranges for sheep on every ration; the only group difference was a slightly higher bilirubin reading in the 15% group at the end of the trial, still inside the physiological range. Lambs on the 15% ration scored higher on body condition at mid-trial and at the end. The study measured neither carcass quality nor cannabinoid residues in tissue, ran for a single season on one breed and one sex, and used ten animals per group, so it establishes that hemp seed meal did no harm at these inclusion levels rather than proving it delivers a production advantage. The findings sit alongside a ten-week European perch trial published earlier in 2026, in which 20% hemp meal improved growth and feed utilisation. In the European Union hemp seed and hemp seed cake are already listed feed materials for all species; in the United States the only hemp feed ingredient with a tentative AAFCO definition is hemp seed meal for laying hens, capped at 20% of the diet.
Why Hemp Seed Meal Keeps Showing Up in Feed Research
A hemp seed is roughly a quarter to a third oil and a fifth to a third protein, with fibre and carbohydrate making up the rest. Pressing or solvent extraction pulls out most of the oil for the food and cosmetics market, and the residue, called hemp seed meal or hemp seed cake, is left with the protein concentrated. The batch used in the Turkish study analysed at 32.19% crude protein and 94.48% dry matter, which puts it in the same league as sunflower meal and not far behind lower-grade soybean meal. The catch is fibre: at 30.41% crude cellulose the meal is far bulkier than soy, which is one reason ruminants, with their four-chambered digestive system built for roughage, are the obvious first customers.
The literature the authors cite reports hemp seed meal ranging from 21.5% to 55.12% crude protein and 8.6% to 11.8% residual fat depending on the seed variety and how hard it was pressed. That variability is itself a finding: a feed formulator cannot treat “hemp seed meal” as a single standard ingredient the way soybean meal is traded, and every batch needs its own analysis. The same point runs through the hemp seed oil composition data we have covered before, where the fatty acid profile shifts noticeably between cultivars.
A 2025 review and meta-analysis in the Journal of Cannabis Research by Agung Irawan and colleagues at Oregon State University pooled 26 feeding studies in ruminants and concluded that hemp by-products generally do not depress performance, while noting that most trials are short and that regulation, not biology, is the main barrier to adoption. That review is available at https://doi.org/10.1186/s42238-025-00279-7. The new lamb study is, in effect, one more data point for that meta-analysis, and a fairly clean one.
The New Lamb Diet Study: 40 Lambs, Four Rations, 75 Days
The design is simple, which is a strength. Forty weaned single male Akkaraman lambs, aged 3.5 to 4 months and weighing 27.63 kg on average with very little spread between them, were bought from one commercial sheep farm and moved to a research facility at 1,050 metres altitude in central Anatolia. Akkaraman is the dominant fat-tailed breed of the Turkish plateau, hardy and slow-maturing, and a very different animal from the Merino-type or Suffolk crosses in most Western feedlot studies. After vaccination, parasite treatment and a two-week adaptation period, the lambs were randomly split into four groups of ten and housed in individual 1.5-by-1.5-metre pens in a semi-open shelter from June to August 2025.
The four rations were built to be identical in energy (2,700 kcal of metabolisable energy per kilogram) and protein (16% crude protein). Hemp seed meal went in at 5%, 10% or 15% of dry matter, and to keep the protein constant the sunflower seed meal was cut from 13% in the control ration to 12%, 6% and 4%, and cottonseed meal from 10% to 7.5%, 7.5% and 6%. Barley stayed at 40% and maize around 24 to 26% in every ration. The concentrate-to-roughage ratio was 80:20, with alfalfa hay as the roughage, a standard intensive fattening setup. Feed was weighed in and weighed back every morning for every lamb so that individual intake could be calculated, and the animals were weighed every 15 days after an overnight fast.
Blood was drawn from the jugular vein at the start and end of the trial and analysed for a panel of 14 markers covering energy metabolism (glucose, non-esterified fatty acids, beta-hydroxybutyrate), protein metabolism (urea, total protein, albumin), minerals (calcium, magnesium, zinc) and liver function (bilirubin, AST, ALP, GGT). Body condition was scored by three assessors independently, and nine body measurements from wither height to chest girth were taken by one person at the beginning, middle and end.
Growth and Feed Efficiency: Where the Numbers Landed
The headline is that nothing moved. Live weight at every 15-day weigh-in was statistically identical across the four groups. By day 75 the control lambs weighed 46.17 kg, the 5% group 46.37 kg, the 10% group 47.23 kg and the 15% group 47.43 kg, with a p-value of 0.740, which means the small upward drift with more hemp is well within what random variation would produce. Total gain over the trial was 19.11 kg, 19.64 kg, 19.60 kg and 19.80 kg respectively, again with no significant difference. Averaged over the 75 days, that works out to roughly a quarter of a kilo per lamb per day on all four rations.
There was one period where the groups diverged. Between days 31 and 45 the control lambs gained 4.46 kg while the 5% hemp group gained 3.00 kg, and the difference was statistically significant. But the 10% and 15% groups sat in between and were not different from either, and the gap closed completely in the following periods. The authors report it without over-interpreting it, and that is the right call: with ten animals per group, a single 15-day window is exactly where noise shows up.
Feed intake showed some scattered significant differences in individual periods. In the second fortnight the 15% group ate noticeably more than the control, for instance, and in the fourth fortnight the 5% group ate slightly less. Over the whole trial, however, intake did not differ, and, more importantly, feed conversion ratio, the kilograms of dry matter needed to put on one kilogram of live weight, did not differ in any period or overall. That is the number a farmer cares about. Hemp seed meal at up to 15% of the ration neither improved nor worsened how efficiently these lambs turned feed into body mass.
Body measurements told the same story: no group differences at any point. Body condition score, a hands-on assessment of fat cover, did differ. At mid-trial the 15% group scored 2.94 against 2.73 to 2.77 for the others, and at the end it scored 3.52 against 3.21 for the control. The authors do not offer a mechanism, and it would be a stretch to read a production benefit into a subjective score when the objective weights and measurements show none.
Blood Chemistry and Liver Health
Blood results are where a feed trial reveals whether an ingredient is quietly doing damage even when the animals look fine. Here the answer was reassuring. Across the 14 markers, the only between-group difference at the end of the trial was bilirubin, which was higher in the 15% group. The authors note that this remained within the normal range for sheep and that a similar bilirubin rise was seen by Nathan Parker and colleagues at Oregon State University in a 2022 study of lambs fed spent hemp biomass, published in the Journal of Animal Science at https://doi.org/10.1093/jas/skac263, also without any sign of liver dysfunction.
The three markers veterinarians use to flag a serious energy deficit all stayed comfortably in the safe zone. Non-esterified fatty acids never reached the 0.8 mmol/L threshold, beta-hydroxybutyrate never reached 1 mmol/L and glucose never fell below 40 mg/dL, in any group at any time point. Liver enzymes AST, ALP and GGT rose between the start and end of the trial in every group, hemp or no hemp, which the authors attribute to normal growth and increased bone and muscle activity in young animals rather than to the diet. Magnesium and cholesterol also shifted between the two sampling points in the hemp groups, plausibly reflecting the meal’s mineral and fat content, but again without leaving the normal range.
What the blood panel does not include is any measurement of cannabinoids. That omission matters and we return to it below.
The European Perch Trial: Where This Story Started
Finding sustainable ways to feed the world’s growing fish population is a major challenge for modern aquaculture. Traditional fish meal is becoming increasingly expensive and environmentally taxing due to overfishing and climate change. However, a study published earlier this year in the journal Animals suggests that industrial hemp (Cannabis sativa) could be part of a greener future for fish farming.
Researchers tested various levels of hemp meal as a protein substitute for European perch, a popular freshwater species. For ten weeks, fish were fed diets containing up to 30% hemp meal to see if it could replace traditional animal protein without harming their health. The results were remarkably positive, showing that a 20% inclusion of hemp meal actually boosted growth performance and feed utilisation.
The study found that hemp meal did not negatively affect the fish’s physical condition, muscle composition or blood parameters. More importantly, the fish receiving hemp showed beneficial changes in their intestinal lining, which can lead to better digestion and nutrient absorption. This suggests that hemp isn’t just a “filler” ingredient but a functional component that supports the overall homeostasis of the digestive system.
Beyond the biology, using hemp in aquafeed makes sense for the planet. Industrial hemp is known for its high CO2-binding capacity and low pesticide requirements, making it a much more eco-friendly crop than many alternatives. By reducing the industry’s dependence on wild-caught fish meal, farmers can lower costs while promoting a circular economy.
Put the two trials side by side and a pattern emerges. In perch, 20% hemp meal was better than the control; in lambs, up to 15% was equal to it. Neither trial found a level at which hemp seed meal hurt the animal. The difference in outcome is probably less about hemp than about what it replaced: the perch study displaced expensive animal protein, the lamb study displaced other plant meals that were already doing the job adequately. Hemp’s advantage in ruminant rations, if it has one, will come from price and local availability rather than from any nutritional magic. The same logic runs through the Moroccan work on cannabis in poultry feed, which was driven as much by the search for antibiotic alternatives as by protein.
Seed Meal Is Not the Whole Plant: The THC Question
The most important distinction in this entire field is between feeds made from hemp seed and feeds made from the rest of the plant. Cannabinoids, including THC, are produced in the resin glands on the flowers and, to a lesser extent, the leaves. The seed itself contains essentially none; whatever THC turns up in seed products is surface contamination from contact with flower material during harvest and cleaning. Hemp seed meal is therefore the lowest-cannabinoid fraction of the plant a farmer can buy.
That distinction is the whole reason the cautionary studies do not contradict the reassuring ones. In 2022 a team at Germany’s Federal Institute for Risk Assessment led by Bettina Wagner fed dairy cows a silage made from hemp leaves, flowers and seeds together, and found that THC transferred into milk at a rate of about 0.2% of what the cows ate. Because cows eat a great deal, that was enough to push the THC in milk above the acute reference dose for some groups of consumers, and the cows themselves became drowsy, unsteady and ate less. That paper, in Nature Food at https://doi.org/10.1038/s43016-022-00623-7, is frequently cited as evidence against hemp in feed. It is evidence against whole-plant hemp silage with a high cannabinoid load, which is a different ingredient with a different risk profile. We have seen the same distinction play out in research on hemp leaves in dog food, where the leaf fraction carried measurable cannabinoids that the seed fraction did not.
The lamb study did not measure cannabinoids in feed, blood or tissue, and that is its most significant gap from a food-safety perspective. The authors bought the meal from a commercial supplier and did not report its THC content. The animals showed no behavioural signs of cannabinoid exposure and their blood chemistry was normal, which is consistent with a low-THC seed product, but “consistent with” is not the same as “measured”. Anyone using this study to argue for hemp seed meal in meat-producing animals should be honest that residue data in lamb meat and fat are still missing from it.
The Regulatory Picture: Europe and the United States
In the European Union the question of whether hemp seed meal may be fed to sheep was settled in principle 15 years ago. The European Food Safety Authority’s animal feed panel published a scientific opinion in 2011, available at https://doi.org/10.2903/j.efsa.2011.2011, concluding that hemp seed, hemp seed cake and hemp seed oil could be used as feed materials for all animal species, and recommending a maximum THC content of 10 mg/kg for hemp seed-derived feed materials. The same opinion recommended that whole-plant hemp feed materials be placed on the list of restricted feeds, drawing precisely the seed-versus-plant line described above. Hemp seed expeller and hemp seed meal appear in the EU Catalogue of Feed Materials as ordinary oilseed by-products. A 2015 EFSA opinion on THC in milk went further and concluded that milk from cows fed hemp seed-derived materials at reported concentrations is unlikely to pose a health concern. For a Turkish or European sheep farmer, then, the legal path to using the meal in the new study’s rations is already open, and the practical questions are price, supply and batch analysis.
The United States is a different story. Hemp was removed from the Controlled Substances Act by the 2018 Farm Bill, but feed ingredients are approved separately, through the Food and Drug Administration’s Center for Veterinary Medicine and the Association of American Feed Control Officials, whose model definitions most states adopt. On 7 August 2024 AAFCO members voted to accept a tentative definition, T71.5, for “Hemp Seed Meal, Mechanical Extracted”, as reported at the time by Marijuana Moment. It is the first hemp-derived ingredient to get that far. It is also narrow: the definition covers laying hens only, at no more than 20% of the diet, with limits of 2 ppm THC and 20 ppm total CBD in the meal. No hemp ingredient is currently defined for sheep, cattle, goats or pigs in the US, and the FDA’s position remains that hemp is not approved for use in animal food beyond that one tentative definition. A US sheep producer reading the Turkish study cannot lawfully replicate it in a commercial flock, even though the meal itself is legal to grow and press.
The gap between those two regimes is not really about the science, which is broadly the same on both continents. It is about who bears the burden of proof and how much residue data regulators want to see before meat and milk from hemp-fed animals enter the food chain. That is exactly why the missing cannabinoid measurements in the lamb study matter more in Washington than in Brussels. Europe’s own regulatory attention to hemp products is currently focused elsewhere, on the EFSA safety assessment of CBD as a novel food, which concerns human rather than animal consumption but shapes how cautiously the whole hemp category is handled.
What the Lamb Study Cannot Tell Us
It is worth listing the limits plainly, because the conclusion “hemp seed meal is safe for lambs up to 15%” is going to be quoted more widely than the caveats. The trial lasted 75 days, a normal finishing period but not a lifetime. It used one breed, one sex and one age band, on one farm in one summer. Ten animals per group is adequate for detecting large effects and inadequate for detecting small ones, so “no significant difference” here means “no large difference”, not “identical”. The lambs were not slaughtered as part of the study, so there are no carcass weights, no meat quality measurements and no fatty acid profiles of the meat, all of which are exactly the data that would tell us whether hemp changes the product rather than just the animal. No cannabinoid residues were measured anywhere. And the hemp seed meal came from a single commercial batch whose composition, as the authors’ own literature review shows, may not match the next batch.
None of this makes the study weak. It is a competent, well-controlled feeding trial with individual-pen intake measurement, which is more than many published feed studies manage, and it was reviewed and accepted within four months of submission. It does what it set out to do: show that a common Turkish meat breed tolerates hemp seed meal at practical inclusion rates without any measurable cost. The authors themselves end by calling for longer trials in other species. The honest summary is that the evidence base for hemp seed meal in ruminants has moved from “probably fine” to “fine in every trial so far”, while the evidence base for what ends up on the plate is still thin.
Frequently Asked Questions
No. Final live weights on the 0%, 5%, 10% and 15% hemp rations were 46.17, 46.37, 47.23 and 47.43 kg, and the differences were not statistically significant. Total gain, daily gain and feed conversion ratio were also the same across groups. The lambs on the 15% ration scored higher on body condition, but their weights and body measurements did not differ from the others. The study shows hemp seed meal can replace part of the sunflower and cottonseed meal without a penalty, not that it delivers a bonus.
The perch trial found that 20% hemp meal actually improved growth and feed utilisation over the control diet, while the lamb trial found equivalence rather than improvement at up to 15%. The most likely explanation is what hemp replaced: expensive animal protein in the fish diet versus other plant meals in the lamb diet. Both trials agree that hemp seed meal at these levels did not harm health, growth or blood chemistry.
Hemp seeds contain essentially no cannabinoids; any THC in seed meal comes from contamination by flower material during harvest. This is why regulators treat seed-derived feeds very differently from whole-plant hemp silage, which was shown in a 2022 German dairy study to transfer THC into milk. The lamb study did not measure cannabinoids in the feed, blood or tissue, so it cannot answer the residue question directly. The lambs showed no behavioural or blood-chemistry signs of cannabinoid exposure.
Not under current rules. In August 2024 the Association of American Feed Control Officials accepted a tentative definition for mechanically extracted hemp seed meal, but it covers laying hens only, at up to 20% of the diet and with THC capped at 2 ppm. No hemp ingredient is currently defined for sheep, cattle, goats or pigs, and the FDA’s Center for Veterinary Medicine has not approved hemp for those species. In the European Union, by contrast, hemp seed and hemp seed cake have been recognised feed materials for all species since EFSA’s 2011 opinion, subject to a recommended 10 mg/kg THC limit. Feed rules are set at national and state level and change; producers should confirm the current position with their feed regulator before formulating.
Legal Disclaimer
This article is a journalistic summary of published scientific research and regulatory documents intended for general information. It is not veterinary, agricultural, nutritional or legal advice. The findings described relate to a specific breed, age group, inclusion rate and feeding period under controlled conditions and may not apply to other animals or production systems. Rules governing the use of hemp-derived materials in animal feed differ between countries, and within the United States between states, and are subject to change. Readers considering hemp seed meal in any commercial ration should consult a qualified animal nutritionist and confirm the applicable feed regulations with the competent authority in their jurisdiction. The Cannex does not endorse any feed product and has no commercial relationship with the researchers, institutions or suppliers mentioned.