Anxiety herb evidence guide · 20-source ledger

Best Herbs for Anxiety: What Human Evidence Actually Supports

Last evidence review August 22, 2026

Anxiety is one of the easiest supplement topics to overstate. A positive stress study is not the same as a generalized-anxiety-disorder trial, and a studied proprietary extract is not the same as the whole plant category. This guide ranks the evidence by diagnostic directness, preparation match, replication, funding concentration and safety — and includes negative trials alongside positive ones.

Anxiety-related botanicals arranged for an evidence comparison
Evidence belongs to the preparation that was studied. Oral lavender oil, lavender tea and aromatherapy are not interchangeable interventions.

Bottom line

Silexan has the most direct anxiety-disorder evidence here — but it comes with a concentration caveat

Five placebo-controlled trials involving 1,213 adult outpatients found oral Silexan 80 mg/day superior to placebo across investigator- and patient-rated anxiety outcomes over ten weeks.[1] A large GAD trial also tested 80 and 160 mg/day against placebo and paroxetine.[2]

That is unusually direct evidence for an herbal product, but it is not independent replication across unrelated lavender preparations. The 2023 meta-analysis states that the included studies were completed by the manufacturer, the work/publication received manufacturer financial support, one author was a company employee and several authors disclosed company ties.[1]

Ashwagandha has repeated-dose stress/anxiety evidence but greater population and formulation heterogeneity.[6][7] Chamomile has a smaller but legitimate GAD research program.[16][17][18] Passionflower remains preliminary, while kava combines mixed efficacy with a materially higher safety concern.[8][10][14]

Directness before popularity

Evidence-ranked anxiety herb comparison

OptionMost direct evidenceMain limitationVerdict
Silexan oral lavender oilFive placebo-controlled anxiety trials / 1,213 adults plus GAD active-comparator work.[1][2]Proprietary preparation; evidence is concentrated around one manufacturer and associated investigators.Most direct trial program
AshwagandhaMultiple RCTs/meta-analyses reporting stress/anxiety improvements.[6][7]High heterogeneity; many populations are stress-enriched rather than one diagnosed anxiety disorder; extracts differ.Promising, less diagnostically direct
Chamomile extractPlacebo-controlled GAD trial plus longer continuation study and meta-analysis.[16][17][18]Small evidence base; long-term study primary relapse endpoint was not significant.Interesting second-tier evidence
Passionflower36-person GAD pilot versus oxazepam.[8]Far too small to establish equivalence; NCCIH says conclusions remain limited.[9]Preliminary
KavaOlder mixed trials, one positive 2013 GAD study and a negative larger 2020 GAD trial.[10][11][12]Inconsistent efficacy plus rare severe liver injury reports.[13][14]Risk-benefit ceiling is high

A broader 2022 network meta-analysis of medicinal herbs concluded that the field remains preliminary because of small samples and strong placebo effects.[19] A systematic review focused specifically on GAD likewise found most complementary approaches supported by limited evidence and poor safety reporting.[15]

1. Oral lavender oil (Silexan): strongest directness, narrowest generalization

The 2023 meta-analysis included all five completed double-blind placebo-controlled trials of Silexan 80 mg/day in adults with subthreshold anxiety, mixed anxiety/depressive disorder or GAD and found statistically significant benefits over placebo after ten weeks.[1] Separate studies include a 539-person GAD trial, a 221-person subsyndromal-anxiety placebo trial and a lorazepam comparison.[2][3][4] A second meta-analysis found improvements in somatic-anxiety symptoms as well.[5]

Funding/directness caveat: the evidence is unusually coherent partly because it studies one defined product. That is scientifically useful, but independence is limited: manufacturer sponsorship and author/company ties are disclosed in the meta-analysis.[1]

Do not convert this into “lavender works for anxiety.” The defensible claim is about a specific oral lavender-oil preparation studied at defined doses and durations.

2. Ashwagandha: meaningful signal, but broader and messier evidence

A 2024 meta-analysis of nine RCTs (558 participants) reported improvements in perceived stress, Hamilton Anxiety scores and cortisol versus placebo.[6] A 2022 analysis of 12 trials (1,002 participants) likewise reported stress/anxiety improvements but had very high heterogeneity and rated certainty low.[7]

That makes ashwagandha credible enough to discuss, but it is less direct for diagnosed anxiety than a GAD-specific trial program. Results also belong to specific extracts and repeated-dose schedules rather than every ashwagandha product or same-day use.

NCCIH says some preparations may help stress but describes anxiety evidence as unclear; it also flags short-term-only safety knowledge, pregnancy/breastfeeding avoidance, thyroid/autoimmune cautions, medication interactions and rare liver injury reports.[20]

Ashwagandha evidence guide →

3. Chamomile: real GAD trials, including an important non-significant endpoint

A placebo-controlled trial in 57 randomized adults with mild-to-moderate GAD found a greater reduction in Hamilton Anxiety scores with pharmaceutical-grade chamomile extract over eight weeks.[16]

The longer-term continuation study is more nuanced. Responders to open-label chamomile were randomized to continue chamomile or switch to placebo. The primary time-to-relapse result favored chamomile numerically but was not statistically significant, although anxiety symptoms remained lower in the continuation group.[17] A 2019 meta-analysis found a GAD signal but little evidence for generic state anxiety and called for larger trials.[18]

Verdict: more interesting than a generic “calming tea” reputation suggests, but still too small a literature for a confident top ranking.

4. Passionflower: the famous oxazepam comparison is only a pilot

The study most often cited enrolled 36 GAD outpatients for four weeks and compared passionflower extract with oxazepam. It reported no significant difference at the end of treatment and less job-performance impairment with passionflower.[8]

But a tiny active-comparator study that fails to detect a difference is not proof of equivalence. NCCIH still describes the overall passionflower anxiety evidence as limited and notes drowsiness, dizziness, confusion, pregnancy and perioperative cautions.[9]

Passionflower evidence profile →

5. Kava: mixed efficacy, unusually important safety tradeoff

Kava illustrates why evidence reviews should include later negative trials. A 2013 six-week GAD study reported a moderate benefit over placebo,[11] but pooled earlier GAD trials found no effect,[12] and a larger 16-week study in 171 adults found no significant advantage over placebo; liver-function abnormalities were also more frequent with kava in that trial.[10]

A 2018 systematic review found mixed short-term evidence, with only three of seven placebo-controlled trials positive.[13] NCCIH notes rare severe and sometimes fatal liver injury reports and warns against combining kava with other sedatives such as benzodiazepines or alcohol.[14]

Verdict: even if efficacy were clearer, the safety ceiling would keep kava from being an easy “best herb” recommendation.

Safety can outrank efficacy

  • • Kava: rare severe liver injury and sedative interactions materially alter its risk-benefit profile.[14]
  • • Passionflower: may cause drowsiness, dizziness or confusion and should not be used during pregnancy; perioperative/anesthesia context matters.[9]
  • • Ashwagandha: pregnancy/breastfeeding, thyroid/autoimmune disorders, liver history and medication interactions deserve explicit review.[20]
  • • Silexan: evidence applies to the manufactured oral preparation, not ingestion of raw lavender essential oil.

What not to infer from these studies

  • • “No significant difference” in a tiny pilot does not prove two treatments are equally effective.
  • • A proprietary extract trial does not validate every product made from the same plant.
  • • A stress score improvement in otherwise healthy adults is not identical to treating diagnosed GAD.
  • • Several positive studies from one manufacturer are not the same evidentiary situation as independent replication by unrelated groups.
  • • A same-day subjective calming effect does not establish durable treatment of an anxiety disorder.

Source ledger

References

20 sources

  1. 01
    Efficacy of Silexan in patients with anxiety disorders: a meta-analysis of randomized, placebo-controlled trials Dold M, et al. Eur Arch Psychiatry Clin Neurosci. 2023;273(7):1615-1628. Five trials / 1,213 adult outpatients.
  2. 02
    Lavender oil preparation Silexan is effective in generalized anxiety disorder: randomized comparison to placebo and paroxetine Kasper S, et al. Int J Neuropsychopharmacol. 2014;17(6):859-869. 539 adults with GAD.
  3. 03
    Silexan is effective in subsyndromal anxiety disorder: a randomized, double-blind, placebo-controlled trial Kasper S, et al. 221 adults, 80 mg/day for 10 weeks.
  4. 04
    A multi-center, double-blind, randomised study of Silexan in comparison to Lorazepam for generalized anxiety disorder Six-week active-comparator GAD trial.
  5. 05
    Therapeutic effects of Silexan on somatic symptoms and physical health in patients with anxiety disorders: A meta-analysis Meta-analysis of five randomized placebo-controlled anxiety trials.
  6. 06
    Effects of Ashwagandha on stress and anxiety: A systematic review and meta-analysis Arumugam V, et al. Explore (NY). 2024. Nine RCTs / 558 participants.
  7. 07
    Does Ashwagandha supplementation have a beneficial effect on the management of anxiety and stress? Systematic review and dose-response meta-analysis of 12 RCTs / 1,002 participants; evidence certainty rated low.
  8. 08
    Passionflower in the treatment of generalized anxiety: a pilot double-blind randomized controlled trial with oxazepam Akhondzadeh S, et al. J Clin Pharm Ther. 2001;26(5):363-367. 36 outpatients.
  9. 09
    Passionflower: Usefulness and Safety National Center for Complementary and Integrative Health. Current evidence and safety overview.
  10. 10
    Kava for generalised anxiety disorder: A 16-week double-blind, randomised, placebo-controlled study Sarris J, et al. Aust N Z J Psychiatry. 2020;54(3):288-297. 171 adults with GAD; no significant benefit over placebo.
  11. 11
    Kava in the treatment of generalized anxiety disorder: a double-blind, randomized, placebo-controlled study Sarris J, et al. J Clin Psychopharmacol. 2013;33(5):643-648. Shorter GAD trial with positive anxiety signal.
  12. 12
    Kava in generalized anxiety disorder: three placebo-controlled trials Connor KM, Payne V, Davidson JRT. Int Clin Psychopharmacol. 2006. Pooled small trials found no kava benefit.
  13. 13
    The effectiveness and safety of Kava Kava for treating anxiety symptoms: A systematic review and analysis of randomized clinical trials Smith K, Leiras C. Complement Ther Clin Pract. 2018;33:107-117. Mixed short-term trial evidence.
  14. 14
    Kava: Usefulness and Safety National Center for Complementary and Integrative Health. Current liver-injury and sedative-interaction warnings.
  15. 15
    Complementary and Alternative Medicine Treatments for Generalized Anxiety Disorder Systematic review and meta-analysis of randomized controlled trials. Evidence for most CAM approaches remained limited.
  16. 16
    A randomized, double-blind, placebo-controlled trial of oral chamomile extract therapy for generalized anxiety disorder Amsterdam JD, et al. J Clin Psychopharmacol. 2009;29:378-382. 57 randomized adults with mild-to-moderate GAD.
  17. 17
    Long-term chamomile treatment for generalized anxiety disorder: A randomized clinical trial Mao JJ, et al. Phytomedicine. 2016;23(14):1735-1742. Primary relapse endpoint was not statistically significant.
  18. 18
    Therapeutic efficacy and safety of chamomile for state anxiety, generalized anxiety disorder, insomnia, and sleep quality Systematic review and meta-analysis of 12 randomized/quasi-randomized trials.
  19. 19
    Medicinal herbs for the treatment of anxiety: A systematic review and network meta-analysis Network meta-analysis concluding herbal results remain preliminary because of small samples and strong placebo effects.
  20. 20
    Ashwagandha: Usefulness and Safety National Center for Complementary and Integrative Health. Current safety review: short-term use, pregnancy/breastfeeding, thyroid/autoimmune, medication and rare liver-injury cautions.

Frequently asked questions

Which herb has the strongest direct evidence for anxiety?

Among the options compared here, oral Silexan has the most concentrated anxiety-disorder trial program, including placebo-controlled GAD and subthreshold-anxiety studies. But the claim belongs to a proprietary oral lavender-oil preparation, not lavender in general, and the 2023 meta-analysis reports substantial manufacturer involvement and author ties that should temper confidence.

Does ashwagandha help anxiety?

Meta-analyses report reductions in anxiety and stress measures with specific ashwagandha preparations, but trial heterogeneity is high and one larger review rated the certainty of evidence low. Many studies also enroll people with elevated stress rather than a single diagnosed anxiety disorder.

Does chamomile help generalized anxiety disorder?

Small controlled studies suggest a possible benefit. One placebo-controlled trial found a greater reduction in anxiety scores with chamomile, while a longer continuation study did not significantly reduce the primary relapse endpoint despite lower symptom scores. A 2019 meta-analysis described the GAD signal as promising but called for larger trials.

Is passionflower as effective as oxazepam?

That has not been established. The often-cited study included only 36 GAD outpatients over four weeks. Finding no significant difference in a small pilot is not proof of equivalence or non-inferiority. NCCIH continues to describe passionflower anxiety evidence as limited.

Does kava work for anxiety?

The evidence is mixed. Some older and shorter trials were positive, but pooled small studies and a larger 16-week GAD trial did not show a significant benefit over placebo. Kava also carries a rare but serious liver-injury concern, which materially changes its risk-benefit profile.

Is lavender aromatherapy the same as Silexan?

No. The strongest direct anxiety evidence on this page is for a defined orally administered lavender-oil preparation. It should not be generalized to lavender tea, aromatherapy, topical products, culinary lavender, or ingestion of raw essential oil.

Can herbs replace therapy or prescribed anxiety medication?

No. Persistent or impairing anxiety deserves evidence-based mental-health care. Do not stop or reduce prescribed treatment because of a supplement guide, and review herb-drug interactions with a clinician or pharmacist.

Free safety checklist

Get the evidence-first supplement notes

Occasional research updates, practical safety context, and new guide announcements. No diagnosis, treatment, or personal medical advice.

We use your email to send the checklist and occasional evidence-first supplement updates. Unsubscribe anytime. Privacy policy.

Newsletter

Get the evidence without the hype

Research updates on herbs, supplements and safety—written around what the human evidence actually supports.

Open newsletter