Herbs for Anxiety: 10 Herbal Remedies Ranked by What the Trials Found
The herbs with the best evidence for anxiety are ashwagandha, kava and lavender oil (the standardised extract called Silexan), followed by passionflower and chamomile. All five have been tested in randomised controlled trials against placebo, and two of them (kava and Silexan) have been compared head to head with prescription anxiolytics. Most other herbs sold for anxiety have either one small trial or none, so this guide ranks every pick by how much human evidence actually exists and tells you the studied dose rather than the label dose.
This page is for adults with everyday or mild-to-moderate anxiety: the racing mind, the tight chest before meetings, the 3 a.m. worry loop. Herbs can take the edge off that, and a few of them do so measurably on validated scales such as the Hamilton Anxiety Rating Scale (HAM-A). They are not a substitute for treatment if you have panic attacks, anxiety that stops you working or leaving the house, thoughts of self-harm, or anxiety alongside heavy drinking. In those cases see a doctor first; CBT and SSRIs have far larger evidence bases than anything below, and a few herbs here interact with those drugs.
One honest caveat before the list. Most herbal anxiety trials are 4 to 12 weeks long, funded by the manufacturer of the extract tested, and run in a few hundred people at most. 'Strong' on this page means several RCTs or a meta-analysis point the same way; it does not mean the effect is large. Expect a modest, real benefit, not a switch.
Quick answer
- Ashwagandha — lowers anxiety scores and morning cortisol in at least five placebo-controlled RCTs at 300–600 mg/day
- Kava — the only herb shown to beat placebo for generalised anxiety disorder in a Cochrane review; liver cautions apply
- Lavender oil (Silexan, 80 mg) — matched low-dose paroxetine and lorazepam for GAD in German trials
- Passionflower — comparable to oxazepam in one small trial, with less next-day drowsiness
- Chamomile — reduced GAD symptoms over 8 weeks and cut relapse in a 26-week follow-up study
- L-theanine — calms the acute stress response within an hour; useful for situational nerves rather than a disorder
The 10 best herbs and supplements for anxiety
Ranked by the strength of human evidence. Strong = multiple trials or a meta-analysis. Moderate = a few small trials. Preliminary = early or animal data.
In a 60-day RCT of 64 adults with chronic stress, 300 mg of root extract twice daily cut Perceived Stress Scale scores by 44% versus 5.5% on placebo and lowered serum cortisol by 28%. A 2014 systematic review of five trials found consistent reductions in anxiety scores, and later trials (250 mg and 600 mg/day, 8 weeks) reproduced the cortisol and HAM-A improvements. It is the best-replicated herb on this list, though most trials are small and industry-funded.
- Key evidence
- 2012 RCT (Chandrasekhar) and 2014 systematic review of 5 RCTs (Pratte)
- Studied dose
- 300–600 mg/day of a root extract standardised to withanolides (KSM-66 or Sensoril), for 6–8 weeks
- Best for
- Anxiety that comes with burnout, poor sleep and feeling wired-but-tired
- Cautions
- May raise thyroid hormone levels, so avoid with hyperthyroidism or if you take levothyroxine without monitoring. Rare cases of liver injury are reported; stop if you notice dark urine or jaundice. Avoid in pregnancy and stop 2 weeks before surgery because of its sedative effect.
A Cochrane review of 12 placebo-controlled trials found kava extract reduced HAM-A scores significantly more than placebo, and a 6-week Australian RCT in 75 people with generalised anxiety disorder found a 26% greater drop in HAM-A on 120–240 mg kavalactones daily, with 26% of the kava group reaching remission versus 6% on placebo. Effects are noticeable within 1 to 2 weeks, faster than most herbs. The catch is a small but real risk of liver injury, which led to bans in several countries in the 2000s.
- Key evidence
- Cochrane review (Pittler and Ernst, 2003) and 2013 GAD RCT (Sarris)
- Studied dose
- 120–250 mg kavalactones/day from a water-based noble-root extract, split in 2–3 doses, for no more than 8 weeks without a break
- Best for
- Short courses for generalised anxiety when you need something that works within days
- Cautions
- Do not combine with alcohol, benzodiazepines, or any sedating drug, and avoid entirely if you have liver disease or take hepatotoxic medication (including high-dose paracetamol). Check liver enzymes if you use it for more than a few weeks. Not for pregnancy or breastfeeding.
Silexan, an 80 mg oral capsule of standardised lavender oil, has been tested in more than 10 RCTs in Germany. In a 10-week trial of 539 adults with GAD it reduced HAM-A scores by 14.1 points at 160 mg versus 9.5 on placebo and performed as well as 20 mg paroxetine. An earlier 6-week trial found it matched 0.5 mg lorazepam. Benefits appear by week 2 and it is not sedating or habit-forming. The evidence is for the oral capsule, not for tea or aromatherapy, which is why it ranks here and not lower.
- Key evidence
- 2014 multicentre RCT vs paroxetine (Kasper), 539 participants
- Studied dose
- 80–160 mg/day of oral Silexan (Lasea or CalmAid) for 6–10 weeks
- Best for
- Generalised anxiety with restlessness and disturbed sleep, when you want a non-sedating option
- Cautions
- The main side effect is lavender-flavoured burping, which most people tolerate. No serious interactions are reported, but it has not been tested alongside SSRIs or in pregnancy, so do not assume safety there. Topical and inhaled lavender have far weaker evidence.
A 4-week double-blind trial in 36 people with GAD found passionflower extract (45 drops/day) lowered HAM-A scores as much as 30 mg oxazepam, with slower onset but less impairment of job performance. Two further trials found a single dose reduced pre-operative anxiety as well as midazolam or oxazepam without the sedation. The total number of participants across all trials is under 200, which keeps it at Moderate despite the encouraging comparisons.
- Key evidence
- 2001 RCT vs oxazepam (Akhondzadeh), 36 participants
- Studied dose
- 250–500 mg/day of a dried aerial-part extract, or 45 drops of tincture daily; 1–2 g as tea at night
- Best for
- Anxious people who also struggle to fall asleep
- Cautions
- Additive sedation with benzodiazepines, sleeping tablets, alcohol and antihistamines. Avoid in pregnancy because of reports of uterine stimulation, and stop 2 weeks before surgery. Rare reports of irregular heart rhythm at high doses.
An 8-week placebo-controlled trial of 57 people with mild-to-moderate GAD found a 220–1,100 mg/day chamomile extract produced a significantly greater drop in HAM-A than placebo. A larger 2016 study treated 179 people openly for 12 weeks and then randomised responders to continue or switch to placebo for 26 weeks: anxiety scores stayed lower on chamomile, though relapse rates did not differ significantly. Mild effect, excellent safety record, and the only herb here where a cup of tea is a reasonable starting dose.
- Key evidence
- 2009 RCT (Amsterdam) and 2016 long-term study (Mao), 179 participants
- Studied dose
- 500–1,500 mg/day of a German chamomile extract standardised to 1.2% apigenin, or 2–3 strong cups of tea daily, for 8 weeks
- Best for
- Mild, chronic worry; people who want the gentlest option
- Cautions
- Avoid if you are allergic to ragweed, daisies or chrysanthemums. Chamomile has mild blood-thinning activity, so mention it to your doctor if you take warfarin. May add to the effect of sedatives.
A 15-day open study of 20 stressed volunteers given 600 mg/day of a standardised lemon balm extract (Cyracos) reported a 15–18% drop in anxiety-related symptoms and insomnia. Earlier double-blind crossover trials in healthy adults found single doses of 300–600 mg improved calmness and reduced the anxiety induced by a laboratory stress test, with mild sedation at the top dose. It also shows up in combination products with valerian that have beaten placebo for stress-induced sleep problems. Promising, but no placebo-controlled trial in a clinical anxiety population yet.
- Key evidence
- 2011 open-label trial of Cyracos (Cases) and 2004 crossover stress study (Kennedy)
- Studied dose
- 300–600 mg/day of a leaf extract standardised to rosmarinic acid, or 1.5–4.5 g dried leaf as tea
- Best for
- Daytime jitteriness and stress-related indigestion
- Cautions
- May interact with thyroid medication by reducing TSH activity; avoid with hypothyroidism unless supervised. Additive sedation with sleeping tablets and alcohol. Generally safe, but little pregnancy data.
L-theanine
Moderate evidenceThe green-tea amino acid reduces the heart-rate and salivary IgA response to an acute mental stressor within 30–60 minutes of a 200 mg dose, and a 2019 RCT in 30 healthy adults found 200 mg/day for 4 weeks lowered stress-related symptoms and improved sleep quality versus placebo. Trials in people with diagnosed anxiety disorders are mixed: one 8-week study in GAD found no benefit over placebo as an add-on to antidepressants. Best thought of as a situational calmer rather than a treatment.
- Key evidence
- 2019 placebo-controlled RCT (Hidese) and 2007 acute stress study (Kimura)
- Studied dose
- 200 mg as needed before a stressful event, or 200–400 mg/day for 4 weeks
- Best for
- Performance anxiety, exam nerves, and caffeine jitters
- Cautions
- Very few side effects. May lower blood pressure slightly, so monitor if you take antihypertensives. Not well studied in pregnancy.
Magnesium glycinate
Moderate evidenceA 2017 systematic review of 18 studies found magnesium supplementation reduced subjective anxiety in people vulnerable to it, including those with mild anxiety and premenstrual anxiety, but rated the trial quality as poor. The effect is most plausible in people who are low in magnesium, which is common on Western diets. Glycinate is the form chosen here because it is well absorbed and the glycine itself has mild calming and sleep-promoting effects.
- Key evidence
- 2017 systematic review of 18 studies (Boyle)
- Studied dose
- 200–400 mg elemental magnesium/day as glycinate, with the evening meal, for 6–8 weeks
- Best for
- Anxiety with muscle tension, cramps, poor sleep or a low-magnesium diet
- Cautions
- Reduce the dose if stools loosen. Separate from antibiotics (tetracyclines, quinolones) and bisphosphonates by 2–4 hours. Check with a doctor first if you have kidney disease, since magnesium can accumulate.
A 10-week open study of 80 people with mild anxiety found 400 mg/day of a standardised rhodiola extract reduced anxiety, stress and anger scores, and an earlier pilot RCT in 10 people with GAD showed a 10-point HAM-A drop on 340 mg/day. Its better evidence is for stress-related fatigue and burnout, where a 28-day placebo-controlled trial found reduced cortisol awakening response and improved mental performance. Mildly stimulating, so it suits the exhausted-anxious rather than the wired-anxious.
- Key evidence
- 2015 open-label trial in mild anxiety (Cropley), 80 participants
- Studied dose
- 200–400 mg/day of a root extract standardised to 3% rosavins and 1% salidroside, taken in the morning
- Best for
- Anxiety mixed with exhaustion, low mood and poor concentration
- Cautions
- Can cause jitteriness or insomnia if taken late in the day. Theoretical interaction with SSRIs and MAOIs (serotonin effects) and with blood-pressure and diabetes drugs. Avoid in bipolar disorder because of reports of triggering agitation.
A 6-week double-blind RCT of 150 adults with symptomatic stress found 1,200 mg/day of a holy basil extract improved a composite of stress symptoms (forgetfulness, exhaustion, sleep problems) about 1.6 times more than placebo. A smaller trial in 35 people with GAD reported reduced anxiety and depression scores over 60 days, but it had no placebo arm. A 2017 review found all 24 human trials reported a benefit, which is itself a sign of publication bias as much as efficacy.
- Key evidence
- 2017 systematic review of 24 human trials (Jamshidi and Cohen)
- Studied dose
- 500–1,200 mg/day of a leaf extract, or 2–3 cups of tulsi tea, for 6–8 weeks
- Best for
- General stress load with blood-sugar swings and fatigue
- Cautions
- Lowers blood glucose, so monitor if you take diabetes medication. Mild antiplatelet effect; stop 2 weeks before surgery and discuss with your doctor if on anticoagulants. Avoid in pregnancy (animal data suggest it can affect fertility).
How to choose
- If you want the most evidence per capsule and your anxiety comes with poor sleep or burnout, start with ashwagandha at 300 mg twice daily and give it 6 weeks.
- If you need relief within days for generalised anxiety and have a healthy liver and no alcohol habit, kava works fastest; keep it to a 4–8 week course and get liver enzymes checked if you continue.
- If sedation is a problem (you drive, you work shifts), choose Silexan lavender oil or L-theanine, neither of which impairs alertness in trials.
- If sleep is the main problem and the anxiety follows from it, passionflower or chamomile in the evening is a better fit than a daytime adaptogen; see our sleep guide.
- If you are exhausted as well as anxious, rhodiola in the morning is the only pick here with a mild lift; avoid it if you are already wired.
- Do not stack three or four of these at once. Pick one, run it for the trial duration, and judge it against a scale such as GAD-7 before adding anything.
What to skip
Valerian for daytime anxiety
Valerian has reasonable sleep evidence but the two placebo-controlled anxiety trials (one in GAD, one in pre-surgical anxiety) found no significant effect, and it is drowsy-making during the day. Keep it for bedtime.
St John's wort
It has decent evidence for mild depression, but trials in anxiety disorders are negative or missing, and it induces liver enzymes that lower the levels of SSRIs, the contraceptive pill, warfarin and many other drugs. Too many interactions for too little anxiety benefit.
CBD gummies and oils
The positive trials used 300–600 mg of pharmaceutical-grade CBD, roughly 10 to 50 times the dose in a typical gummy. At shop doses the evidence is essentially nil, and product testing repeatedly finds the labelled CBD content is wrong.
Oral GABA
GABA barely crosses the blood-brain barrier, and the handful of small trials showing relaxation effects measured brain waves rather than anxiety. If something in a 'GABA' product works, it is usually the L-theanine or magnesium alongside it.
Safety and when to see a clinician
- See a clinician first if you have panic attacks, anxiety that prevents normal life, thoughts of self-harm, chest pain or palpitations with anxiety, or if symptoms started suddenly after age 40 (thyroid, heart and medication causes need ruling out).
- Sedative herbs (kava, passionflower, chamomile, lemon balm) add to benzodiazepines, z-drugs, opioids, antihistamines and alcohol. Never combine them with these without medical advice, and do not drive until you know how you respond.
- Serotonergic overlap: rhodiola and (in theory) ashwagandha may interact with SSRIs, SNRIs and MAOIs. Tell your prescriber before adding either; do not stop an antidepressant to start a herb.
- Thyroid: ashwagandha can push thyroid hormones up and lemon balm can push them down. Both need monitoring if you take levothyroxine or have a thyroid condition.
- Liver: kava carries a documented, rare risk of liver injury and ashwagandha has case reports. Stop and get a blood test if you develop nausea, dark urine, pale stools or yellowing skin.
- Pregnancy and surgery: none of the herbs above has adequate pregnancy safety data, and most should be stopped 2 weeks before surgery because of sedative or blood-thinning effects.
What matters as much as any herb
- Cut caffeine to under 200 mg/day (about two small coffees) or stop it for two weeks as a test: caffeine reliably provokes anxiety symptoms in sensitive people, and no herb out-performs removing the trigger.
- Aerobic exercise three times a week reduces anxiety symptoms with an effect size comparable to many herb trials, and it works in people with diagnosed anxiety disorders, not just the stressed-well.
- Protect 7 to 9 hours of sleep: a single night of short sleep increases next-day anxiety by around 30% in laboratory studies, and most herbal anxiety benefit in trials came with better sleep.
- Alcohol relieves anxiety for about two hours and raises it for the next 24; if you drink to calm down, that habit is likely doing more harm than any supplement can offset.
- A self-help CBT programme or app (the evidence base is large) addresses the thinking patterns that keep anxiety going. Herbs lower the volume; CBT changes the song.
Frequently asked questions
What is the best herb for anxiety?
By weight of evidence, ashwagandha for chronic stress-type anxiety and kava or Silexan lavender oil for generalised anxiety disorder. Ashwagandha has the most replicated trials and the cleanest safety profile for medium-term use; kava works fastest but needs liver precautions; Silexan is the only one tested head to head against an SSRI and a benzodiazepine.
How long do herbs for anxiety take to work?
Kava and L-theanine act within hours to days. Ashwagandha, chamomile, lavender oil and passionflower showed significant differences from placebo at 2 to 4 weeks in trials, with full effect by 6 to 8 weeks. If a herb has done nothing after 8 weeks at a studied dose, it is not going to.
Can I take herbs for anxiety with antidepressants?
Some combinations are probably fine (chamomile, L-theanine, magnesium) and some are not (St John's wort, which reduces SSRI levels; rhodiola, which may add serotonergic effects). Kava and passionflower add sedation to mirtazapine and trazodone. Always tell your prescriber, and never swap a herb in for a prescribed antidepressant on your own.
Which herbal tea is best for anxiety?
Chamomile has the only clinical-dose evidence that translates to tea: 2 to 3 strong cups a day gets you close to the lower doses used in trials. Lemon balm and passionflower teas are reasonable evening choices. Lavender tea is pleasant but not the form that was tested; the anxiety evidence is for the 80 mg oral oil capsule.
Is ashwagandha or kava better for anxiety?
Kava produces a bigger and faster drop in anxiety scores in generalised anxiety disorder, but it should be limited to short courses because of liver risk and it interacts with alcohol and sedatives. Ashwagandha is slower and milder but can be taken for months and also improves sleep and cortisol. For most people reading this, ashwagandha is the safer first choice.
Are herbal sedatives for anxiety addictive?
None of the herbs on this list produces the tolerance and withdrawal seen with benzodiazepines, and the long-term chamomile and ashwagandha studies found no rebound when they were stopped. Kava has been used habitually in Pacific cultures without dependence, though heavy use causes a reversible skin condition. The addiction risk is low; the interaction risk with real sedatives is the thing to watch.
What herbs should I avoid for anxiety?
Stimulating herbs make anxiety worse: guarana, yerba mate, high-dose panax ginseng, ephedra and anything containing synephrine or large amounts of green tea extract. Yohimbe provokes panic in susceptible people and is used in research to induce anxiety. If you are already on a sedative, avoid kava and valerian unless a doctor agrees.
References
- Chandrasekhar K, Kapoor J, Anishetty S. A prospective, randomized double-blind, placebo-controlled study of safety and efficacy of a high-concentration full-spectrum extract of ashwagandha root in reducing stress and anxiety in adults. Indian Journal of Psychological Medicine (2012)View source
- Pratte MA, Nanavati KB, Young V, Morley CP. An alternative treatment for anxiety: a systematic review of human trial results reported for the Ayurvedic herb ashwagandha (Withania somnifera). Journal of Alternative and Complementary Medicine (2014)View source
- Sarris J, Stough C, Bousman CA, et al.. Kava in the treatment of generalized anxiety disorder: a double-blind, randomized, placebo-controlled study. Journal of Clinical Psychopharmacology (2013)View source
- Pittler MH, Ernst E. Kava extract for treating anxiety. Cochrane Database of Systematic Reviews (2003)
- Kasper S, Gastpar M, Müller WE, et al.. Lavender oil preparation Silexan is effective in generalized anxiety disorder: a randomized, double-blind comparison to placebo and paroxetine. International Journal of Neuropsychopharmacology (2014)View source
- Akhondzadeh S, Naghavi HR, Vazirian M, et al.. Passionflower in the treatment of generalized anxiety: a pilot double-blind randomized controlled trial with oxazepam. Journal of Clinical Pharmacy and Therapeutics (2001)View source
- Amsterdam JD, Li Y, Soeller I, et al.. A randomized, double-blind, placebo-controlled trial of oral Matricaria recutita (chamomile) extract therapy for generalized anxiety disorder. Journal of Clinical Psychopharmacology (2009)View source
- Mao JJ, Xie SX, Keefe JR, et al.. Long-term chamomile (Matricaria chamomilla L.) treatment for generalized anxiety disorder: a randomized clinical trial. Phytomedicine (2016)View source
- Boyle NB, Lawton C, Dye L. The effects of magnesium supplementation on subjective anxiety and stress: a systematic review. Nutrients (2017)View source
- Hidese S, Ogawa S, Ota M, et al.. Effects of L-theanine administration on stress-related symptoms and cognitive functions in healthy adults: a randomized controlled trial. Nutrients (2019)View source
- Kava. National Center for Complementary and Integrative Health (2020)View source
- Lopresti AL, Smith SJ, Malvi H, Kodgule R. An investigation into the stress-relieving and pharmacological actions of an ashwagandha (Withania somnifera) extract: a randomized, double-blind, placebo-controlled study. Medicine (Baltimore) (2019)View source
Related conditions
More guides
Not sure which one fits you?
Ask AskMN. It knows these studies, checks interactions with the medications you tell it about, and it is free and unlimited, no account needed.
Ask about anxietyThis guide is educational and is not medical advice. Herbs and supplements can interact with prescription medicines and are not regulated like drugs. Talk with a clinician or pharmacist before starting anything new, especially if you are pregnant, breastfeeding, have a chronic condition, or take medication.