Benefits
Anxiety and mood rating scores alongside a prescribed antidepressant
In a single-blind trial in 60 adults with diagnosed depression, all taking the antidepressant sertraline, a nightly leaf extract syrup added for 4 weeks lowered Hamilton anxiety and Beck depression scores more than a placebo syrup. It was an add-on, not a substitute for prescribed treatment. Basil's taste may have revealed group assignment, and the groups differed in age.
Sleep quality and sleeplessness scores in women around menopause
In a triple-blind trial in 60 Iranian women aged 40 to 65, one capsule of hydroalcoholic leaf extract (250 mg) each night for a month gave better sleep quality and insomnia severity scores than placebo at 2 and 4 weeks. Scores were self-reported, and no repeat trial was found.
Menopausal symptom ratings after one month of leaf extract
In the same trial, total Menopause Rating Scale scores did not differ from placebo at 2 weeks but were lower than placebo after 1 month, as were the psychological and somatic subscales; the urogenital subscale did not differ. Basil users started with higher scores, some table values do not match the text, and ratings were self-reported.
Mood and sleep scores in the weeks after childbirth
In a triple-blind trial in 78 Iranian women without depression at enrollment, two 500 mg capsules of dried basil leaf and aerial part powder daily, started within 72 hours of birth and taken for 8 weeks, did not lower postnatal depression scores or improve sleep quality compared with starch capsules. The authors suggested testing an extract instead.
Body weight and body fat with basil seed fiber taken before meals
In a 12-week trial in 120 adults with fatty liver, 10 g a day of basil seeds stirred into water before lunch and dinner did not change weight, BMI or body fat compared with an untreated control, although weight fell within groups. Across the four arms (psyllium included), 50 people left; 8 could not tolerate the seed or psyllium drinks.
Hemoglobin and iron markers with basil leaf powder cookies
In a 4-month trial in 204 adolescent girls in Pakistan with low iron status, cookies made with 16% dried basil leaf powder ended with higher hemoglobin and serum iron than plain cookies, similar to iron-fortified cookies. The hemoglobin values in the text and tables do not match, so the size of the effect is unclear.
Mechanism of action
Aroma compounds depend on the variety
Basil oil is a blend of linalool, eugenol, methyl eugenol and other compounds whose mix depends on the chemotype. In a study of 21 cultivars, estragole was absent from Genovese and purple basil types but present in Napoletano types, so the estragole in a basil product depends on the plant used.
Sleep-lengthening effects of injected extract in mice
In mice, basil leaf extract injected into the abdomen lengthened sleep induced by the sedative pentobarbital and shortened the time to fall asleep. This was an injection, not an oral dose, and it does not show how basil acts in people.
A basil flavonoid slows estragole activation in rats
Estragole becomes reactive in the liver after enzymes convert it to a sulfate metabolite. The basil flavonoid nevadensin blocked this step in lab tests and cut liver DNA adducts by 36% in rats dosed with both. Modeling suggests this effect may be small at human intake levels, and the EMA does not count it toward a safe threshold.
Seed mucilage swells into a gel
Basil seeds are coated with a mucilage that absorbs water and becomes thick and sticky. In the seed studies, seeds were taken with a glass of water or swollen in about 240 mL of water before meals.
Clinical trials
Randomized, single-blind, placebo-controlled trial at one psychiatric clinic in Fasa, Iran, of a hydroalcoholic basil leaf extract syrup (1,100 mg extract in 5 mL, nightly) added to sertraline 50 to 100 mg/day in adults with major depressive disorder; no specific funding. (Talaei et al. 2025, Brain Behav)
60 adults randomized, 57 analyzed (27 basil, 30 placebo); the basil group was younger on average (34 vs 40 years).
Hamilton Anxiety scores fell by 20.4 points on basil versus 9.7 on placebo, and Beck Depression scores by 18.4 versus 10.7 (both P < 0.001 between groups). All 3 dropouts were in the basil group (1 for stomach discomfort) and were not analyzed. Mild side effects on basil were stomach discomfort (1) and headache (2). The authors note basil's taste and smell may have unblinded participants. Single small trial; basil was an add-on to sertraline, not a replacement.
Triple-blind, randomized, placebo-controlled trial in Mashhad, Iran (IRCT20200104046001N1) of one capsule nightly holding 250 mg hydroalcoholic basil leaf extract plus 250 mg filler, versus starch capsules; sleep results and menopausal symptom results (Eur J Med Res 2024) were published separately; no specific funding. (Karimi et al. 2023, Phytother Res)
76 menopausal women aged 40 to 65 enrolled; 16 (8 per group) were excluded during the study, leaving 60 analyzed (30 per group).
At 1 month, Pittsburgh sleep quality scores were 3.7 on basil vs 9.1 on placebo and insomnia severity 5.6 vs 11.0 (both significant). Total Menopause Rating Scale scores did not differ at 2 weeks (9.5 vs 11.2, P = 0.163) but did at 1 month (6.9 vs 11.3, P = 0.001); the somatic subscale differed only at 1 month in the paper's table, although its abstract says both time points. Basil users started with higher symptom scores (12.7 vs 10.8), and several table values do not match the text. No serious side effects were reported. Self-rated scales; one city.
Triple-blind, randomized, placebo-controlled trial at Taleghani Hospital, Tabriz, Iran, of two 500 mg capsules of dried basil leaf and aerial part powder daily versus two 500 mg starch capsules (placebo bottles scented with basil oil), starting within 72 hours of birth, to see if basil lowers postnatal depression scores; funded by Tabriz University of Medical Sciences. (Abdi et al. 2025, Heliyon)
78 postpartum women with Edinburgh Postnatal Depression Scale scores of 12 or less at entry; 77 analyzed.
At 8 weeks, median depression scores were 1.0 in both groups (P = 0.966); sleep quality scores (scale given at week 6 per the methods) were 17.63 vs 16.15 (P = 0.400). The primary and secondary outcomes were both null. One participant reported nausea. The authors suggested higher doses or an extract for future work.
Four-arm, randomized trial in Isfahan, Iran, of basil seeds 10 g/day, psyllium 10 g/day, a mix of both, or an untreated control (no placebo), taken in two 5 g doses with water before lunch and dinner; funded by Isfahan University of Medical Sciences. (Akbarian et al. 2016, Int J Prev Med)
120 adults with nonalcoholic fatty liver disease randomized; 8 could not tolerate the seed or psyllium drinks and 42 more left, across all four arms.
Weight and BMI fell within groups, but changes in weight, BMI, body fat and lean mass did not differ between groups after 12 weeks. Participants could not be blinded because they stirred the seeds into water. Reasons for intolerance were constipation, nausea or stomach ache.
Randomized controlled trial in Sargodha, Pakistan, of cookies with 16 g basil leaf powder per 100 g, cookies fortified with ferrous sulfate, or plain control cookies, given as five cookies (50 g) designed to supply half the daily iron need; no funding reported. (Akbar et al. 2024, Food Sci Nutr)
204 adolescent girls with iron deficiency anemia (68 per group).
Hemoglobin, serum iron and other blood markers rose in the basil and iron cookie groups and fell in the plain cookie group, with significant differences between groups. The hemoglobin values given in the text and in the tables do not match, and blinding is not described, so effect sizes are uncertain.
Animal study of a hydroalcoholic basil extract and its fractions (25 to 100 mg/kg) injected into the abdomen of mice 30 minutes before pentobarbital, compared with saline and diazepam, plus cell toxicity tests. (Askari et al. 2016, Iran Red Crescent Med J)
72 mice in 9 groups; neuronal PC12 and L929 fibroblast cell lines.
The extract lengthened pentobarbital-induced sleep at all three doses, comparably to diazepam, and shortened sleep latency. It did not harm the cell lines, and the LD50 of the injected extract was 2.4 g/kg. Injection in mice is not comparable to oral use in people.
Rat study with physiologically based kinetic modeling, in which estragole and the basil flavonoid nevadensin were given orally together at a ratio found in basil; author affiliations include flavor industry bodies and Nestlé. (Alhusainy et al. 2013, Mol Nutr Food Res)
Male Sprague-Dawley rats; computer models of rat and human liver.
Adding nevadensin cut estragole DNA adducts in rat liver by 36%. The model predicted stronger inhibition in human liver. A related study predicted the effect may be limited at lower, realistic doses, and the EMA judged such matrix effects not yet enough to set a safe threshold.