Basil: nobody eats it by the hundred grams

A meta-analysis pooling 45 interventional studies in adults with type 2 diabetes looked for blood sugar effects across twelve Mediterranean herbs and spices. It found them in cinnamon, turmeric, ginger, black cumin and saffron, with the largest falls in fasting glucose, 27 to 17 mg/dL, under black cumin, cinnamon and ginger. Basil was not among the herbs with an effect. That absence is mostly an absence of trials, which is a different thing from having tested basil and found nothing. The doses in the trials that did work were supplement doses rather than cooking amounts. The second thing to know about this herb is scale. Fresh basil carries 414.8 mcg of vitamin K per 3.5 oz (100 g), which is 346 percent of the Daily Value, and nobody eats 3.5 oz (100 g) of basil. A leaf weighs 0.5 g and a tablespoon of chopped basil 3 g, so a tablespoon works out at about 12 mcg of vitamin K, roughly a tenth of the Daily Value. That 12 mcg is arithmetic across two datasets rather than a portion anyone measured, and the per portion figures in the American survey tables are themselves recalculated from the per 3.5 oz (100 g) value. The one trial that has fed basil to people fed them an extract. Sixty patients at a psychiatric clinic with major depressive disorder and anxiety, all of them already taking 50 to 100 mg of sertraline a day, were given a nightly syrup with 1100 mg of hydroalcoholic basil leaf extract for four weeks. Their Hamilton anxiety scores went from 33.19 to 12.81 against 34.53 to 24.8 on placebo. That is a concentrated extract added to a prescribed drug in one clinic, and this page has no source on what that extract does beyond four weeks or in anyone outside that care. The European Medicines Agency treats estragole, a compound of sweet basil, as a genotoxic carcinogen. For herbal products containing it, EMA asks for intake as low as possible and use for 14 days at most, with a guidance value of 0.05 mg a day. The trial ran four weeks and did not report the estragole content of its extract. That position covers herbal medicinal products, not basil leaves used in cooking.

The numbers, per 3.5 oz (100 g)

As people eat it, per 3.5 oz (100 g)

How much is actually in it

• Fresh basil holds 414.8 mcg of vitamin K and 264 mcg of vitamin A per 3.5 oz (100 g) by the USDA record, while the British dataset gives 660 mcg of vitamin A for the same herb, a 60 percent gap with nothing to settle it. — USDA SR Legacy record for Basil, fresh (FDC 172232) against the CoFID 2021 record for Basil, fresh (13-804) (n = 2 records) Evidence E source
  Vitamin K 414.8 mcg and copper 0.385 mg per 3.5 oz (100 g); vitamin A 264 mcg by USDA against 660 mcg by CoFID, a 60 percent divergence

What a real portion looks like

• Basil was among the five most eaten herbs in a Polish survey of 270 adults, whose whole daily intake of herbs and spices together came to 22.4 g for omnivores and 25.8 g for vegans, with vegans taking in 5.82 mg of salicylates a week against 3.13 mg. — 270 Polish adults aged 19 to 67, 208 omnivores and 62 vegans, reporting habitual use of 61 fresh and dried herbs and spices over the preceding three months (n = 270) Evidence C source
  All herbs and spices together 22.4 plus or minus 18.4 g a day for omnivores against 25.8 plus or minus 25.9 g for vegans, p = 0.007; median salicylate intake 3.13 against 5.82 mg a week, p < 0.001
• A basil leaf weighs 0.5 g and a tablespoon of it weighs 3 g, so a tablespoon carries about 12 mcg of vitamin K, a tenth of the Daily Value, and the 346 percent on the per 3.5 oz (100 g) table describes an amount nobody eats. — household measures recorded for Basil, raw in the food code used for US dietary recalls, read against the USDA vitamin K figure for fresh basil (n = 4 measures) Evidence E source
  1 leaf 0.5 g, 1 tablespoon 3 g; at 414.8 mcg of vitamin K per 3.5 oz (100 g) a 3 g tablespoon gives about 12 mcg against a Daily Value of 120 mcg

What cooking and storage change

• Dried basil shows four times the vitamin K of fresh, 1714.5 against 414.8 mcg per 3.5 oz (100 g), and five times the copper, which is what drying out the water does to any per 3.5 oz (100 g) figure. — the two USDA records for this herb, Basil fresh (FDC 172232) and Spices basil dried (n = 2 records) Evidence E source
  Vitamin K 414.8 mcg fresh against 1714.5 mcg dried, copper 0.385 against 2.1 mg, vitamin A 264 against 37 mcg per 3.5 oz (100 g)

What it does to your health

• Added to sertraline for four weeks in one psychiatric clinic, a nightly syrup carrying 1100 mg of hydroalcoholic basil leaf extract took Hamilton anxiety scores from 33.19 down to 12.81, while placebo went from 34.53 to 24.8. — 60 adults at a psychiatric clinic with major depressive disorder and anxiety, all on sertraline 50 to 100 mg a day, randomized to basil syrup or placebo nightly for 4 weeks, 3 withdrawals from the basil arm (n = 60) Evidence B source
  HAM-A 33.19 to 12.81 with basil against 34.53 to 24.8 with placebo, p < 0.001; Beck depression score 39.44 to 21.04 against 40.3 to 29.57, p < 0.001
• Six grams of a thirteen-spice blend that included basil, stirred into a 1000 calories high-fat meal, lowered the inflammatory signal that blood cells stimulated in a tube released four hours later in 12 men with excess weight, and two grams did not. — 12 nonsmoking men aged 40 to 65 with overweight or obesity and at least one cardiovascular risk factor, each eating the same meal three times with 0, 2 or 6 g of spice blend (n = 12) Evidence B source
  Spice by time interaction on IL-1 beta P < 0.001, IL-8 P = 0.020 and TNF-alpha P = 0.009 from stimulated blood cells, with the IL-1 beta reduction at 240 minutes seen only at the 6 g dose

Safety, limits and interactions

• In market surveys in northeast Thailand, 53.3 percent of Thai basil samples carried intestinal parasites, against 36 percent of all 300 fresh vegetable samples tested. — 300 samples of ten fresh vegetables from 10 markets in five districts of Khon Kaen province, Thailand, each washed in saline and the sediment examined by microscope (n = 300 samples) Evidence C source
  Overall parasite prevalence 36.0 percent, Thai basil and celery 53.3 percent each, peppermint highest at 70.0 percent; Blastocystis hominis 24.7 percent, Strongyloides stercoralis 13.0 percent, Ascaris lumbricoides 8.7 percent
• An outbreak of 46 laboratory-confirmed Shigella sonnei infections in two Norwegian towns was traced to imported basil, with basil pesto the only food that cases in both towns had eaten. — 46 confirmed cases in Tromso and Sarpsborg, Norway, October 2011, investigated by interviews plus a cohort study of banquet attendees and a trace-back of the supply chain (n = 46) Evidence C source
  46 confirmed cases, 42 in one town and 4 in the other, with basil pesto the only common food item and eating basil pesto an independent risk factor at the banquet
• The only dose-finding safety work on linalool, the compound that gives basil its smell, was done in chickens, where feed containing more than 2 percent of it worsened feed conversion and 5 percent enlarged the liver. — day-of-hatch chicks fed diets with graded linalool for 3 weeks, with serum chemistry, gross pathology and relative liver weights measured Evidence D source
  Dose-dependent harm to feed conversion above 2 percent of feed, liver weights significantly increased at 5 percent, aspartate aminotransferase rising with dose but within normal range up to 2 percent
• The European Medicines Agency treats estragole, a compound in sweet basil, as a genotoxic carcinogen and says intake from herbal products should be as low as possible, with use limited to a maximum of 14 days and a guidance value of 0.05 mg a day. — users of herbal medicinal products containing estragole, a compound of sweet basil (Ocimum basilicum) Evidence E source
  Estragole intake from herbal products as low as possible; use limited to a maximum of 14 days; guidance value 0.05 mg per person per day
• In one patient with systemic reactions to oregano and thyme, skin prick tests with extracts of mint-family herbs, basil among them, were all positive, while prick tests with dried basil itself were negative. The authors read this as cross-sensitivity within the family. — one man with systemic allergic reactions to oregano and thyme, tested against other mint-family herbs including basil (n = 1) Evidence D source
  Skin prick tests with mint-family extracts positive in all cases; prick-by-prick with dried basil negative; specific IgE detected to all herbs tested; controls negative

What people believe that the data does not support

• A meta-analysis that searched twelve Mediterranean herbs and spices for effects on blood sugar in type 2 diabetes found them in cinnamon, turmeric, ginger, black cumin and saffron, with the largest falls in fasting glucose, 27 to 17 mg/dL, under black cumin, cinnamon and ginger. Basil was not among the herbs with an effect. — adults with type 2 diabetes in 45 interventional studies pooled out of 77 reviewed, literature searched to September 2023 (n = 45 studies pooled) Evidence A source
  Fasting glucose fell by between 27 and 17 mg/dL with black cumin, cinnamon and ginger; only ginger and black cumin improved HbA1c; only cinnamon and ginger lowered insulin; no glycemic effect reported for basil

Who this works differently for

• For pregnant and breast-feeding women the EMA does not recommend estragole-containing herbal products above 0.05 mg of estragole a day, and for children up to 11 years above 1.0 microgram per kg of body weight. — pregnant and breast-feeding women and children up to 11 years using herbal medicinal products containing estragole Evidence E source
  Not recommended in pregnancy and breast-feeding above 0.05 mg estragole per person per day; not recommended in children up to 11 years above 1.0 mcg/kg body weight
• In the 4-week trial that added 1100 mg of basil leaf extract to sertraline in 60 patients, no serious side effects were seen. In the basil arm one person had passing stomach discomfort and two had mild headaches. — 60 adults at a psychiatric clinic with major depressive disorder and anxiety, all on sertraline 50 to 100 mg a day, basil syrup (1100 mg hydroalcoholic leaf extract) or placebo nightly for 4 weeks (n = 60) Evidence B source
  No serious side effects in either arm; in the basil arm 1 transient gastrointestinal discomfort and 2 mild headaches, all resolving without treatment; 1 withdrawal for gastrointestinal discomfort

The bottom line

Dried basil looks like a different food on the table, with 1714.5 mcg of vitamin K per 3.5 oz (100 g) against 414.8 in the fresh herb and 2.1 mg of copper against 0.385. These are two independent USDA records rather than one batch measured before and after drying, so they are not a retention measurement. Vitamin A moves the other way in them, down from 264 to 37 mcg, which removing water does not explain. The USDA retention table has no group for spices and herbs at all, so real cooking and drying losses for basil have no source here. The two reference datasets also disagree about the fresh herb. USDA gives 264 mcg of vitamin A per 3.5 oz (100 g) and the British CoFID record gives 660 mcg for the same herb, a 60 percent gap. Neither USDA record is a sampled Foundation Foods entry, so that difference cannot be put down to variety, season or method. One human trial has fed basil to people, and it fed them an extract. Sixty adults at a psychiatric clinic had major depressive disorder and anxiety, and all of them were already taking 50 to 100 mg of sertraline a day. For four weeks they were given a nightly syrup carrying 1100 mg of hydroalcoholic basil leaf extract, or a placebo. Hamilton anxiety scores went from 33.19 to 12.81 with the syrup against 34.53 to 24.8 with placebo, p < 0.001. The two groups differed in age at baseline, 34 years against 40, and three people left the basil arm. That is a concentrated extract added to a drug in one clinic, and it says nothing about leaves on a plate. No serious side effects were seen in either arm. In the basil arm one person had passing stomach discomfort and two had mild headaches. With 27 people finishing four weeks on basil, that is far too few to catch a rare interaction with the antidepressant, such as serotonin syndrome, and the trial does not say it looked for one. No study sets a dose limit for the extract itself. The nearest thing is the EMA position on estragole in herbal products: intake as low as possible, 14 days at most, and a guidance value of 0.05 mg a day. In pregnancy and breast-feeding EMA does not recommend such products above 0.05 mg of estragole a day, and in children up to 11 years above 1.0 microgram per kg of body weight. Because the trial extract has no reported estragole content, whether 1100 mg a night stays under that value is unknown. The nearest thing to a culinary dose comes from a blend. Twelve men aged 40 to 65 with excess weight ate the same 1000 calories high-fat meal three times. Six grams of a thirteen-spice blend containing basil lowered the inflammatory signal their blood cells released four hours later, and 2 g did not. Basil is one spice of thirteen there, the cytokines were measured from cells stimulated in a tube, and 6 g of dried spice in one meal is a lot. The other cautions on this page are about the fresh leaf rather than about any dose. An outbreak of 46 laboratory-confirmed Shigella sonnei infections in two Norwegian towns was traced to imported basil, with pesto the only food eaten by cases in both towns. An outbreak names a vehicle rather than a rate. The implicated batch had already left the market when it was identified, so the contamination was never measured, and it was uncooked leaves that carried it. In market surveys in five districts of Khon Kaen province in Thailand, 53.3 percent of Thai basil samples carried intestinal parasites, against 36 percent of all 300 fresh vegetable samples tested. The method was microscopy of the washing sediment, which can over-call and cannot tell living organisms from debris. Finding parasite stages on a leaf is not the same as anyone falling ill, and what both findings support is washing. The only dose-finding work on linalool, the compound that gives basil its smell, was done in chicks, where feed containing more than 2 percent of it worsened feed conversion and 5 percent enlarged the liver. Those are percent-of-diet doses no person eating basil approaches, and they set no human limit. Allergy to basil is barely documented in what we hold. In one man with systemic reactions to oregano and thyme, skin prick tests with extracts of mint-family herbs, basil among them, were positive, while prick tests with dried basil itself were negative. That shows basil can share allergens with its plant family, not how often basil allergy occurs. Anyone taking a vitamin K antagonist is the reader for whom that arithmetic matters. A tablespoon of basil is about 12 mcg against a Daily Value of 120 mcg, and this page has no source on how basil meets the drug. How much herb people use at all is barely measured. In 270 Polish adults reporting their habitual use of 61 herbs and spices over three months, basil was among the five most eaten. All herbs and spices together came to 22.4 g a day for omnivores against 25.8 g for vegans, with salicylate intake at 3.13 against 5.82 mg a week. That is self-reported frequency, the grams are every herb added together rather than basil alone, and no outcome was measured in anyone.

Sources

1. bsl-r1-08 · food composition dataset
   Basil, fresh (FDC 172232): vitamin K 414.8 mcg, vitamin A 264 mcg RAE, copper 0.385 mg per 100 g; CoFID 13-804 Basil, fresh: vitamin A 660 mcg
   USDA SR Legacy, FoodData Central, record 172232 · checked 2026-09-24
2. bsl-r1-07 · cross-sectional survey
   The five most frequently eaten herbs among omnivores were parsley, garlic, dill, marjoram and basil, while among vegans they were garlic, parsley, ginger, basil and dill...Estimated medial salicylates intake was significantly higher among vegans (p = 0.000) and reached 5.82 mg/week vs. 3.13 mg/week for omnivores.
   Nutrients, 2020 · checked 2026-09-24
3. bsl-r1-10 · portion dataset
   Basil, raw (2709780): 1 leaf 0.5 g, 1 tablespoon 3 g
   USDA FNDDS, Basil raw, record 2709780 · checked 2026-09-24
4. bsl-r1-09 · food composition dataset
   Basil, fresh and Spices, basil, dried (USDA SR Legacy): vitamin K 414.8-1714.5 mcg, copper 0.385-2.1 mg, vitamin A 37-264 mcg per 100 g across the 2 records
   USDA SR Legacy, FoodData Central, records 172232 and Spices basil dried · checked 2026-09-24
5. bsl-r1-01 · randomized controlled trial
   After the intervention, the scores were 12.81 ± 4.48 in the basil group and 24.8 ± 5.82 in the placebo group. The reduction in HAM‐A scores was significantly greater in the basil group compared to the placebo group (p < 0.001)...a syrup containing 1100 mg hydroalcoholic extract per 5 mL was prescribed as a single nightly dose to enhance compliance and minimize gastrointestinal discomfort
   Brain and Behavior, 2025 · checked 2026-09-24
6. bsl-r1-03 · randomized crossover
   There was a significant spice-by-time interaction on IL-1β (P < 0.001), IL-8 (P = 0.020), and TNF-α (P = 0.009) secretion from LPS-stimulated PBMCs.
   J Nutr, 2020 · checked 2026-09-24
7. bsl-r1-04 · prevalence survey
   The overall prevalence of IPs was found to be 36.0%, with Blastocystis hominis (24.7%), Strongyloides stercoralis (13.0%), and Ascaris lumbricoides (8.7%) being the most common. Peppermint showed the highest prevalence rate at 70.0%, followed by celery and Thai basil at 53.3% each.
   Am J Trop Med Hyg, 2025 · checked 2026-09-24
8. bsl-r1-05 · outbreak cohort investigation
   In total, 46 cases were confirmed (Tromsø= 42; Sarpsborg= 4)...Basil pesto was the only common food item that had been consumed by confirmed cases occurring in Tromsø and Sarpsborg.
   Euro Surveill, 2013 · checked 2026-09-24
9. bsl-r1-06 · animal feeding study
   Linalool had a dramatic negative dose-dependent effect on feed conversion at concentrations in the feed exceeding 2% linalool, but not on gross pathology. Liver weights were significantly increased in the 5% linalool-treated birds.
   Poult Sci, 2014 · checked 2026-09-24
10. hf2-basil-1 · agency position
    It is concluded that the intake of estragole from HMPs in the general population should be as low as practically achievable because of the generally accepted evidence of genotoxic carcinogenicity. ... It is concluded, that the intake of estragole from HMPs in the general population should be as low as possible, which includes a short-time duration of use (maximum 14 days) and a discussion about the single / daily doses necessary for adults and adolescents according to the risk assessment relevant for the concerned HMP. For example, to reach or come as close as possible to the guidance value of 0.05 mg/person per day, the lowest dose should be consistently selected if ranges of single and daily doses are available from traditional use.
    European Medicines Agency, HMPC, Public statement on the use of herbal medicinal products containing estragole, EMA/HMPC/137212/2005 Rev 1 Corr 1 (2023) · checked 2026-10-08
11. hf2-basil-4 · case report
    Skin tests with plants of the Labiatae family were positive in all cases when the skin prick technique was used; tests were negative with basil and lavender, and positive with all the others when we used the prick by prick technique. ... Specific IgE was detected to all herbs tested; higher levels were obtained with the CAP system. CONCLUSIONS: Plants belonging to the Labiatae family seem to show cross-sensitivity on the basis of clinical history and in vitro and in vivo test results.
    Ann Allergy Asthma Immunol, 1996 · checked 2026-10-08
12. bsl-r1-02 · meta-analysis
    Our results showed that cinnamon, turmeric, ginger, black cumin, and saffron significantly improved the fasting glucose levels in T2DM subjects. The most significant decreases in fasting glucose were achieved after supplementation with black cumin, followed by cinnamon and ginger, which achieved a decrease of between 27 and 17 mg/dL.
    Nutrients, 2024 · checked 2026-09-24
13. hf2-basil-2 · agency position
    The usage of estragole containing HMPs in pregnant and breast-feeding women is not recommended if the daily intake of estragole exceeds the guidance value of 0.05 mg/person per day, unless otherwise justified by a risk assessment based on adequate safety data. ... The usage of estragole containing HMPs in children up to 11 years is not recommended if the daily intake of estragole exceeds the guidance value of 1.0 µg/kg bw, unless otherwise justified by a risk assessment based on adequate safety data.
    European Medicines Agency, HMPC, Public statement on the use of herbal medicinal products containing estragole, EMA/HMPC/137212/2005 Rev 1 Corr 1 (2023) · checked 2026-10-08
14. hf2-basil-3 · randomized controlled trial
    Both the basil and placebo supplements were well‐tolerated, and no serious side effects were observed during the study period. Mild side effects in the basil group included transient gastrointestinal discomfort (n = 1) and mild headache (n = 2), all resolving without intervention.
    Brain and Behavior, 2025 · checked 2026-10-08