- Kava (Piper methysticum) is a Pacific pepper-family root. Its active compounds, kavalactones, have a calming effect. The US National Center for Complementary and Integrative Health (NCCIH) says kava "may be helpful for anxiety" but "does not appear to be helpful" for generalised anxiety disorder (NCCIH, April 2025).
- The Cochrane review pooled 7 small trials (380 people) and found kava lowered Hamilton anxiety scores by 3.9 points more than placebo (95% CI 0.1 to 7.7). The review judged the effect small and not robust (Pittler & Ernst, Cochrane; full text).
- The largest and longest trial, a 16-week Australian study in 171 people with generalised anxiety disorder, found kava no better than placebo. Remission was 17.4% on kava vs 23.8% on placebo, and abnormal liver tests were more frequent on kava (Sarris et al. 2020).
- Kava is a recognised, if rare, cause of serious liver injury. LiverTox counts 50 to 100 published cases and gives kava its highest likelihood score, "A" (LiverTox).
- Rules differ sharply by country. Kava food and medicines are banned in England (SI 2002/3169; SI 2002/3170). The EU herbal committee refused to write a kava monograph (EMA HMPC). Australia allows a cold-water drink made from "noble" kava root, with drowsiness warnings (FSANZ 2025).
- Money trail: The 2020 trial was funded by Australia's National Health and Medical Research Council and co-funded by Integria Healthcare (MediHerb), which made the tablets (trial protocol). The earlier positive 2013 trial lists Integria as its funder and primary sponsor (ANZCTR). Several older positive trials used the WS 1490 extract, which the trial reports name; the Cochrane product breakdown and the extract's maker could not be verified from accessible sources.
- Evidence grade: Weak for short-term relief of anxiety symptoms. Insufficient for sleep and one-off "situational" use. Moderate evidence that it does not treat diagnosed generalised anxiety disorder. Risk for the liver, sedative interactions and driving.
Independent evidence review · Supplement
Kava probably has a small calming effect over a few weeks, but the best independent evidence says it does not treat generalised anxiety disorder, and it carries a rare but real risk of serious liver injury. Its positive record rests on small older trials, several of them of the WS 1490 extract, and on Australian trials linked to a kava-tablet maker. The largest and longest trial, co-funded by that kind of maker and by Australia's NHMRC, was negative. Regulators in Europe and the UK have not licensed kava as an anxiety medicine, while Pacific nations and Australia treat traditional water-based kava as a food with warnings. Our confidence: moderate that it fails for diagnosed anxiety disorders; low for any other benefit.
Table of contents
- Evidence summary
- What kava is, and its legal status
- How it works
- What works and what does not
- Benefits by claim
- Risks and all side effects
- All interactions
- Who should avoid kava
- Doses studied and product quality
- Follow the money: who makes it and who funded the evidence
- Related research
- Frequently asked questions
- Sources and funding notes
Evidence summary
| Claim | Evidence | Source | Funding / conflict | Strength |
|---|---|---|---|---|
| Kava extract reduces anxiety symptoms vs placebo | Twelve trials; seven pooled (n=380). HAM-A difference 3.9 points (95% CI 0.1 to 7.7; p=0.05). The authors say the effect "seems small" and "lacks robustness". | Pittler & Ernst, Cochrane | Cochrane author affiliations and declarations not re-checked (full text paywalled). Included trials' funders not checked. | Weak |
| Kava treats diagnosed generalised anxiety disorder (GAD) | 16 weeks, n=171: 1.37-point difference favouring placebo (p=0.25); remission 17.4% kava vs 23.8% placebo. Three earlier NCCIH-funded US trials (pooled n=64): no effect. | Sarris 2020; Connor 2006 | NHMRC grant co-funded by Integria Healthcare (MediHerb), which made the tablets. Connor: NCCIH grants. | Moderate (against the claim) |
| Earlier Australian trials showed large benefits | 3-week crossover, n=60: effect size d=2.24. 6-week trial, 75 enrolled: remission 26% vs 6% (press-reported). | KADSS 2009; KALM 2013 | 2009: registry lists University of Queensland as funder; a MediHerb co-author link could not be verified. 2013: NHMRC and Integria Healthcare funded it; lead author says the company had no role in design, analysis or publication. | Weak |
| Kava improves sleep in anxious people | One 4-week trial, n=61, of WS 1490: better sleep-quality sub-scores than placebo. | Lehrl 2004 | Extract WS 1490; funding statement not accessed. | Insufficient |
| A single dose calms "situational" anxiety | 22 anxious adults, one dose of 180 mg kavalactones: no change in anxiety; oxazepam reduced it. | Sarris 2012 | Academic (Australia); funding statement not accessed. | Insufficient |
| Kava can cause severe liver injury | LiverTox: 50 to 100 published cases; likelihood score A. WHO 2007 case counts could not be verified. | WHO 2007; LiverTox | WHO-commissioned committee; LiverTox is run by the US NIH | Risk |
| Traditional water-based kava is liver-safe | Hepatitis cases after traditional water-based drinking were reported in New Caledonia (case series). Liver-test abnormalities were also more frequent with a water-based tablet in the 2020 trial. | Russmann 2003; Sarris 2020 | Academic case series; trial as above | Risk (lower) |
| Kava extract causes cancer in animals | "Clear evidence" of liver tumours (hepatoblastoma) in male mice. No IARC classification verified. | NTP TR-571 | US government (NTP) | Risk (animal data) |
| Kava impairs driving | Fiji case-control study: driving within 12 hours of kava linked to 4.7 times the odds of a serious-injury crash (95% CI 1.90 to 11.63). A simulator study of one 180 mg dose found no impairment. | Wainiqolo 2016; Sarris 2013 | Wellcome Trust; simulator study academic | Risk |
Independent evidence and credibility scorecard
Independence tiers: 1 = regulator or government body; 2 = academic or Cochrane team with no declared commercial funding for the work; 3 = independent authors analysing mostly manufacturer-run trials, or minor declared industry ties; 4 = manufacturer-funded or manufacturer-authored. Credibility runs from A (highest) to D.
| Source | Who funds it | Country | Independence tier (1–4) | Credibility (A–D) | Why it would tell the truth / residual bias |
|---|---|---|---|---|---|
| NIH LiverTox: Kava Kava | US National Institute of Diabetes and Digestive and Kidney Diseases (NIH) | US | 1 | A | No product to sell; cites both sides of the causality debate. Last updated April 2018, so newer trials are missing. |
| NCCIH kava fact sheet | US Congress via NIH (appropriations) | US | 1 | A− | Updated April 2025 and reflects the negative 2020 trial. NCCIH also funds research into complementary products, including kava trials. |
| FDA scientific memorandum on kava (2020) | US government | US | 1 | A | A toxicology review written to decide food status, so it focuses on harms rather than benefits. |
| EMA herbal committee (HMPC) public statement | EU agency (public body) | EU | 1 | A | Assessed all products on the EU market and declined a monograph. A short statement with limited detail on efficacy. |
| WHO hepatotoxicity assessment (2007) | WHO; committee commissioned by its medicines-safety advisory committee | International (WHO) | 1 | A− | The most detailed case-by-case causality review. Produced with the Natural Standard Research Collaboration, a commercial evidence database (per its press release). Funding declaration not verified. |
| FSANZ imported-food risk statement (2025) | Australian and New Zealand governments | Australia / NZ | 1 | A | Food regulator with no product to sell. Covers kava as a drink, not tablets or extracts. |
| Pittler & Ernst, Cochrane review | Peninsula Medical School, Universities of Exeter and Plymouth | UK | 2 | B | Declarations not re-checked (full text paywalled). Searches end in January 2005, and 6 of 7 pooled trials used the same company's extract, so the result is close to a review of one product. |
| Connor et al. 2006 (three US trials) | NCCIH grants (grant numbers not verified) | US | 2 | B− | Publicly funded and published a negative result. Very small (64 people). A related NCCIH-registered trial (NCT00083980) was terminated over European liver-toxicity reports; we could not confirm it is one of the three. |
| Sarris et al. 2020 (K-GAD trial) | NHMRC grant APP1063383, co-funded by MediHerb / Integria Healthcare, which manufactured the tablets | Australia | 3 | A− | Large, 16 weeks, pre-registered, and published a result unhelpful to its co-funder. The lead author has declared payments from kava companies. |
| KADSS trial 2009 | Registry lists University of Queensland as funder. A reported MediHerb consultant link could not be verified. | Australia | 3 | C | Very large effect from a 3-week crossover in 60 people. Short and small; the effect was not repeated in the later, larger trial. |
| KALM trial 2013 | Registry lists Integria (commercial) as funder and primary sponsor (ANZCTR) | Australia | 4 | C+ | Diagnosed GAD; 75 enrolled for 6 weeks. The funder makes kava tablets; the lead author says it had no role in design, analysis or publication. |
| WFSBP / CANMAT guideline (2022) | Professional societies; lead author declared industry payments, including from Integria Health, and others | International / Canada | 3 | B | Rated kava "Not Recommended" for GAD despite the lead author's industry ties, which is an admission against interest. |
Note on clustering: most of the modern efficacy trials come from one Australian research group, and several older trials used one extract (WS 1490). The safety evidence is more spread across the US, EU, UK, Australia and WHO.
What kava is, and its legal status
Kava is a shrub in the pepper family, native to the South Pacific. Common names include kava kava, awa, ava and kawa (NCCIH). Pacific communities have drunk it for more than 1,000 years, traditionally by mixing fresh or dried root with cold water in a shared bowl (FSANZ 2025). The active compounds are kavalactones. Six of them make up about 96% of the pharmacologically active content (FAO/WHO 2016).
Forms that matter for safety
- Noble vs non-noble cultivars: There are more than 200 varieties. Traditional drinks use "noble" kava. "Two-day" (tudei) kava is higher in kavalactones and alkaloids and causes nausea and longer-lasting effects (FAO/WHO 2016). Vanuatu's Kava Act 2002 bans the sale or export of tudei and wild kava, with a narrow exception (FSANZ 2025).
- Root vs aerial parts: Kavalactones are highest in the root. The alkaloid pipermethystine, more toxic to liver cells in the laboratory, is found mainly in stems and leaves (FAO/WHO 2016). WHO recommends that only root or rhizome be used, with no aerial parts (WHO 2007).
- Water vs acetone or ethanol extracts: Western tablets were mostly acetone or ethanol extracts. These pull out 2 to 10 times more kavalactones than water and can contain 30–70% kavalactones (FAO/WHO 2016). WHO found more liver problems with organic-solvent extracts than with synthetic products (WHO 2007).
Legal status by country
| Place | Status | Source |
|---|---|---|
| England (UK) | Sale or import of any food containing kava is an offence (in force since 13 January 2003). Kava medicines for human use are prohibited, except for external use. Both were in force as originally made; later amendments were not checked. | SI 2002/3169; SI 2002/3170 |
| European Union | The herbal committee's public statement (21 November 2017) records that no EU herbal monograph was adopted, so no public summary was prepared. Its stated reasons were not verified. | EMA/HMPC/450588/2016 |
| Germany | Authorisations were suspended in 2002 and revoked in December 2007. A Cologne court annulled the revocation in 2014, and the higher court in Münster upheld that in 2015. After the 2015 Münster ruling, products returned to the market on prescription only. A 2019 Cologne ruling, reported in trade press only, set a 200 mg daily kavapyrone limit and fortnightly liver checks. | apotheke adhoc 2015; Pharmazeutische Zeitung 2015; NutraIngredients 2014 |
| United States | FDA issued a consumer advisory on 25 March 2002 (CDC MMWR). Kava is still sold as a supplement. FDA's 2020 toxicology memo is the main federal review of kava as a food ingredient (FDA memo); we did not open it in full. A trade body, the International Kava Organization, says FDA letters in 2025 accepted water-only kava tea as a conventional food; we could not view those letters (IKO press release). | As cited |
| Australia | Food: only a cold-water drink from peeled root of a noble variety, or dried/raw root, labelled "Use in moderation" and "May cause drowsiness" (Standard 2.6.3). Medicines: prescription-only (Schedule 4) unless a registered product is labelled at 250 mg kavalactones a day or less, with 125 mg or less per tablet (Poisons Standard). Commercial imports restarted on 1 December 2021 (FSANZ). | As cited |
| International food standard | Codex regional standard CXS 336R-2020 covers roots, rhizomes or basal stems of noble cultivars for use as a drink mixed with water. | FSANZ 2025 |
| Other markets | FDA's 2020 memo lists withdrawals or bans in France, Switzerland, the Czech Republic, Spain, the UK, Hungary, Portugal, Germany (until 2015) and Canada since 2002. We did not verify each country's current rules. | FDA memo |
How it works
LiverTox states plainly that kava's mechanism "has yet to be elucidated" (LiverTox). What is known:
- Laboratory only: Kavain, a major kavalactone, boosted GABA-A receptor activity in egg-cell (oocyte) experiments, but not through the benzodiazepine site. Effects on NMDA receptors and calcium channels are also proposed (FDA memo).
- In humans: no human study has yet shown how kava relieves anxiety. LiverTox describes its mechanism as not established (LiverTox).
- Sedation and coordination: People intoxicated with kava showed eye-movement, coordination and visual-attention changes, with normal complex thinking (FDA memo). This matters for driving.
- Liver enzymes: In test tubes, kava extract blocked several drug-metabolising enzymes (CYP1A2, 2C9, 2C19, 2D6, 3A4) (Mathews 2002). In 12 volunteers, it inhibited only CYP2E1, by about 40% (Gurley 2005).
What works and what does not
| Claimed benefit | Verdict | Evidence | Key caveat |
|---|---|---|---|
| Short-term relief of anxiety symptoms (mixed groups) | MIXED | HAM-A −3.9 points vs placebo (Cochrane) | Fragile result; small trials; several used the WS 1490 extract |
| Diagnosed generalised anxiety disorder | DOES NOT WORK | No benefit over 16 weeks (Sarris 2020); "Not Recommended" (WFSBP/CANMAT) | One positive 6-week trial was industry-funded |
| Sleep problems linked to anxiety | INSUFFICIENT | One trial, n=61 (Lehrl 2004) | Single extract; not replicated |
| One-off dose for situational anxiety | INSUFFICIENT | No acute effect in 22 adults (Sarris 2012) | NCCIH says effects may take weeks |
| Depression | INSUFFICIENT | Secondary outcome in one 3-week trial (KADSS) | Not a treatment for depression |
Benefits by claim
Does kava work for anxiety?
The Cochrane review included 12 double-blind trials. Seven trials, with 380 people, were pooled on the Hamilton Anxiety scale: kava lowered scores by 3.9 points more than placebo (95% CI 0.1 to 7.7; p=0.05). The authors said the size of the effect seems small (Cochrane). In plain terms, a 3.9-point drop on a 56-point scale is noticeable but modest, and the lower confidence limit is close to zero.
The review's own conclusion is that the effect "seems small" and "lacks robustness", and rests on a relatively small sample (Cochrane).
Generalised anxiety disorder
The strongest test is the K-GAD trial in Australia: 171 unmedicated adults with diagnosed GAD took a water-based noble-kava tablet (240 mg kavalactones a day) or placebo for 16 weeks. Placebo did slightly better (1.37 points; p=0.25), and remission was 17.4% on kava vs 23.8% on placebo (Sarris 2020). The university behind the trial summarised it as "Kava ineffective for Generalised Anxiety Disorder" (Western Sydney University).
Earlier, smaller trials from the same group were positive: a 6-week trial (75 enrolled) reported 26% remission on kava vs 6% on placebo (Sarris 2013), funded by tablet maker Integria according to its registry entry. US trials funded by NCCIH found "no effects" (Connor 2006). A 2018 meta-analysis of three placebo-controlled GAD trials (n=130) found no significant benefit (Ooi 2018). The WFSBP/CANMAT taskforce rated kava "Not Recommended" for GAD (WFSBP/CANMAT 2022).
Everyday stress and "situational" anxiety
Kava's supporters argue it suits short-term, non-clinical anxiety. The evidence for this is thin. A single 180 mg dose had no measurable anti-anxiety effect in 22 anxious adults, while the benzodiazepine oxazepam did (Sarris 2012). NCCIH notes that kava may need "several weeks" to have an effect (NCCIH).
Does kava help you sleep?
One German trial gave 61 people with anxiety-related sleep problems 200 mg of WS 1490 extract or placebo for 4 weeks. Two sleep-questionnaire scores improved more on kava (Lehrl 2004). We could not open the paper's funding statement. A 15-person study of kavalactones for sleep in military trainees was sponsored by supplement maker Thorne HealthTech; its 2024 results paper found mostly null effects on sleep and cortisol (NCT05381025). NCCIH says there is not enough evidence for any use beyond anxiety.
Risks and all side effects
How common is liver injury?
Nobody knows exactly. LiverTox estimates fewer than 1 case per million daily doses, but warns that spontaneous reporting captures under 1% of severe supplement reactions (LiverTox). Trials have not detected cases: seven trials in the Cochrane review measured liver enzymes and found no clinically significant changes (Cochrane). But in the 2020 trial, abnormal liver tests were "significantly more frequent" on kava, though no one met criteria for herb-induced liver injury (Sarris 2020).
The case series
- CDC reported 11 people in Germany, Switzerland and the US who had liver failure and needed transplants after kava (CDC MMWR 2002).
- Traditional water-based kava has also caused hepatitis (Russmann 2003).
Proposed causes include solvent extracts, non-noble cultivars or aerial parts, alcohol, genetic differences in liver enzymes, mould contamination and other medicines (NCCIH; WHO 2007). Aflatoxin, a mould toxin, has been detected in kava root (FSANZ 2025).
Driving and coordination
In a crossover study of 22 adults, one 180 mg dose did not impair simulated driving, while oxazepam slowed braking (Sarris 2013). Real-world drinking amounts are much larger. A 2022 study by Aporosa and colleagues measured cognition during a traditional kava session of about six hours and 3.6 litres per person (20 kava users, 19 controls). It found no significant impairment in five of six measured domains, and a significant regression in one timing-judgement measure in the kava group. The authors say the result needs caution (Aporosa 2022). In Fiji, 23% of drivers in serious-injury crashes had drunk kava in the previous 12 hours vs 4% of comparison drivers: adjusted odds ratio 4.70 (95% CI 1.90 to 11.63) (Wainiqolo 2016).
Kava dermopathy and heavy use
Heavy drinkers can develop a dry, scaly skin rash called kava dermopathy, (Norton & Ruze 1994). NCCIH adds reddened eyes and temporary yellowing of skin, hair and nails. Heavy ongoing use has been linked to this rash and to altered liver tests, which reverse after stopping (FSANZ 2025). In heavy drinkers, GGT (a liver enzyme) rises but ALT usually does not (LiverTox, Clough 2003).
Other side effects
| Effect | What the evidence shows | Source |
|---|---|---|
| Stomach upset, headache, dizziness | The most common side effects. | NCCIH; Sarris 2013 |
| Poorer memory, tremor or shakiness | Each reported by 36 people on kava vs 23 on placebo over 16 weeks. | Sarris 2020 |
| Dependence or withdrawal | No difference from placebo in withdrawal or addiction measures over 6 weeks. LiverTox calls abuse potential "low, but not absent". | Sarris 2013b; LiverTox |
All interactions
| Interacting substance | What can happen | Seriousness | Source |
|---|---|---|---|
| Alcohol | Kava plus alcohol produced larger impairment than alcohol alone. Heavy alcohol intake is a WHO-listed risk factor for liver reactions. | Avoid | Foo & Lemon 1997; WHO 2007 |
| Benzodiazepines (e.g. alprazolam) and other sedatives | Kava is a central nervous system depressant, and combining it with alcohol or benzodiazepines is a recognised interaction (NSW Health). | Avoid | Almeida 1996; NCCIH |
| Medicines that can harm the liver (including paracetamol in excess) | Co-medication with liver-toxic drugs is a WHO risk factor. FSANZ urges caution with paracetamol, opioids and barbiturates. | Avoid or ask a doctor | WHO 2007; FSANZ; FDA memo |
| Drugs cleared by CYP2E1 (e.g. chlorzoxazone) | About 40% inhibition in volunteers. Laboratory inhibition of CYP1A2, 2C9, 2C19, 2D6 and 3A4 was not seen in this human study. | Low–moderate | Gurley 2005; Mathews 2002 |
| Other herbal products | FSANZ advises caution when combined with other herbal preparations. | Caution | FSANZ |
Who should avoid kava
- People with liver disease, or who drink heavily: WHO advises kava should not be used in either group (WHO 2007).
- Pregnancy and breastfeeding: NCCIH warns of "special risks" from "harmful pyrone constituents" (NCCIH). FSANZ found no safety data and says these groups should not drink it (FSANZ).
- Children and adolescents: no safety data (FSANZ). One of the US transplant cases was a 14-year-old (CDC).
- People taking sedatives, antipsychotics, anti-clotting drugs or liver-straining medicines (see interactions).
- People with a diagnosed anxiety disorder: the evidence does not support kava as treatment, and effective options exist.
Doses studied and product quality
These are doses used in research or set by regulators. They are not recommendations.
| Setting | Dose | Source |
|---|---|---|
| Trials reviewed by Ooi 2018 | 120–280 mg kavalactones a day for 4 to 8 weeks | Ooi 2018 |
| K-GAD trial (negative) | 120 mg twice daily (240 mg a day) for 16 weeks; noble Borogu cultivar from Pentecost Island, Vanuatu | Protocol |
| Germany (conditions from 2015 to 2019) | Prescription only; maximum 200 mg kavapyrones a day; up to 2 months' use; liver checks every 2 weeks (2019 Cologne ruling, trade press) | apotheke adhoc |
| Australia (registered medicines) | Labelled maximum 250 mg kavalactones a day; 125 mg or less per tablet or capsule; label warning that it "may harm the liver" | Poisons Standard; TGA |
Product quality problems
- Registered Australian kava medicines must carry a label warning that the product "may harm the liver" (TGA 2022).
Follow the money: who makes it and who funded the evidence
Who grows, makes and sells it
- Pacific producers: Vanuatu, Fiji and Tonga are major kava exporters (export figures not verified).
- Integria Healthcare / MediHerb (Australia): the company named as co-funder and co-sponsor of the K-GAD trial, which also made the tablets (protocol). We could not verify its ownership.
- Fiji Kava Ltd / The Calmer Co (Australia): an ASX-listed seller of kava capsules and powder under the Fiji Kava and Taki Mai brands (FoodNavigator 2023).
- Thorne HealthTech (US): sponsor of a 15-man kava sleep study, published in 2024 in the Journal of the International Society of Sports Nutrition (NCT05381025). Its results were mostly null on sleep and cortisol.
Who funded the trials
- K-GAD (2020): NHMRC project grant APP1063383, co-funded by MediHerb (Integria), which manufactured the kava. The protocol does not describe MediHerb's role in design or analysis (protocol).
- KALM (2013): registry lists Integria, a commercial company, as funding source and primary sponsor (ACTRN12610000381088). The lead author says the company had no role in design, analysis or publication.
- KADSS (2009): the registry entry names the University of Queensland as funder (ACTRN12608000536369). We could not verify the paper's full conflict-of-interest statement.
- Author disclosures: the lead investigator of the Australian trials has declared payments from kava companies including Integria Health, among others (protocol; WFSBP/CANMAT). His group nonetheless published the negative 2020 result.
- Public funders: NCCIH funded the negative US trials (Connor 2006); one NCCIH trial stopped because of "European reports of liver toxicity" (NCT00083980).
- Cochrane: author affiliations and declarations not re-checked this round (full text paywalled).
Documented integrity events
- TGA compliance action (2022): the website for "Noble Kava" (sponsor Fiji Kava (Australia) Trading Pty Ltd) claimed the product treats generalised anxiety disorder and does not harm the liver, although its label had to say it "may harm the liver". The TGA required the sponsor to correct the website; the product was not recalled (TGA).
- Market withdrawals and court reversal: Several countries, including Germany, Switzerland, Canada, Australia and France, restricted kava products after liver reports, according to the CDC's 2002 summary (CDC MMWR 2002). German court rulings of 2014–2015 are covered under legal status.
- US food-status dispute: A federal FDA memo (2020) and a trade body's 2025 press release reach different conclusions on whether kava counts as a food. We did not open either in full (IKO press release).
Research funding at a glance
12 sources graded. The counts below are the independence tiers assigned in the scorecard above. They count sources, not funding amounts or evidence quality.
Public funding alone does not prove independence, and commercial funding alone does not prove a claim false. Named funders, countries and relationships are in the scorecard and source notes.
What this means: The case for kava as an anxiety treatment rests on small older trials, several of one extract, and on Australian trials supported by a kava-tablet maker. The most rigorous of those trials, partly paid for by the same maker, was negative. The case against routine use rests mostly on regulators and public bodies with no product to sell. Pacific economies are heavily reliant on kava trade (the 2025 Economidis review calls it an essential trading commodity), and the newer food standards focus on traditional, water-based noble kava rather than concentrated extracts.
Related research
For wider, evidence-graded context, see these Pure City Research guides:
- Generalised anxiety disorder: treatments, supplement evidence and safety
- Supplements for stress, anxiety and depression: the evidence
- Stress, anxiety and depression: prevention and management guide
- Drug and supplement induced liver injury
- Oral lavender (Silexan)
- Ashwagandha: evidence and safety
- Chamomile for sleep and anxiety
- Valerian
- L-theanine
- Sleep supplements: the evidence
- Buspirone
- Drowsy driving and medicine effects
Frequently asked questions
Does kava actually work for anxiety?
Possibly a little, for a few weeks, in people without a formal diagnosis. Pooled trials showed a 3.9-point drop on a 56-point anxiety scale versus placebo, but the authors describe the effect as small (Cochrane). For diagnosed generalised anxiety disorder, the largest trial found no benefit (Sarris 2020).
Is kava bad for your liver?
It can be, rarely but seriously. LiverTox lists it as a well-known cause of liver injury, including cases needing transplants (LiverTox). Risk seems higher with solvent extracts, high doses, long use, alcohol and other medicines, but water-based kava has also caused hepatitis (WHO 2007; Russmann 2003).
Is kava legal in the UK?
Selling or importing food containing kava is an offence in England (SI 2002/3169), and kava medicines for internal use are prohibited (SI 2002/3170). Both were in force as originally made; later amendments were not checked.
Is kava legal in Australia and the US?
In Australia, a cold-water drink made from peeled noble kava root is a permitted food with warning labels, and registered kava medicines are capped at 250 mg kavalactones a day (FSANZ; Poisons Standard). In the US, kava is sold as a supplement, but its status as a food ingredient is disputed (FDA memo).
Can you drink alcohol with kava?
No. Together they caused more impairment than alcohol alone (Foo & Lemon 1997), and heavy drinking is a risk factor for kava liver injury (WHO 2007). NCCIH advises against combining kava with alcohol.
Can you drive after drinking kava?
It is safest not to. In Fiji, driving within 12 hours of kava was linked to 4.7 times the odds of a serious crash (Wainiqolo 2016). A single tablet dose did not impair simulated driving, but traditional drinking sessions involve far larger amounts (Aporosa 2022).
Does kava help with sleep?
The evidence is one small 4-week trial of one company's extract (Lehrl 2004). NCCIH says there is not enough evidence for uses other than anxiety (NCCIH).
Is kava addictive?
A 6-week trial found no difference from placebo in withdrawal or addiction measures (Sarris 2013). LiverTox describes abuse potential as low but not absent, and Australian authorities cite past kava misuse in some communities (LiverTox; FSANZ).
What is "noble" kava?
Noble varieties are the cultivars traditionally used for drinking in the Pacific. Two-day (tudei) kava is stronger, longer-acting and causes nausea (FAO/WHO 2016). Australian food law and the Codex standard require noble root, but this alone does not remove liver risk (FSANZ).
Sources and funding notes
What we could not verify: The WHO 2007 case counts (the 93-case, 14-transplant and 7-death figures were removed). The Witte 2005 meta-analysis figures (paper not located). The Hawaiian revenue figure. The Cochrane review's declarations and product breakdown. The Cologne 2019 ruling and the KALM 2013 registry entry were not re-opened this round. The EMA statement's stated reasons (PDF text not readable). The KADSS and KALM conflict statements beyond registry records. The Sarris 2012, Lehrl 2004 and Sarris 2013 driving-simulator funding statements. Full texts of the FDA 2020 memo, the NCCIH page, LiverTox, the FSANZ 2025 risk statement and the WHO 2007 report were not opened in full, so claims drawn from them rest on search summaries. The German 2019 ruling is reported in trade press only. Integria Healthcare's ownership is unverified. Germany and KALM dates differ between sources. The LiverTox "transplants and deaths" wording was taken from search summaries, not the full entry. The K-GAD protocol gives an 18-week design; the published trial was 16 weeks.
- LiverTox: Kava Kava (last update 10 April 2018) — liver-injury profile; US NIH (NIDDK).
- NCCIH: Kava: Usefulness and Safety (April 2025) — US government fact sheet; funded by Congress (appropriations history).
- FDA CFSAN scientific memorandum: review of the safety of kava for use in conventional foods (11 August 2020) — US regulator.
- CDC MMWR: Hepatic toxicity possibly associated with kava-containing products (29 November 2002) — US government; records FDA's 25 March 2002 advisory.
- EMA/HMPC public statement on Piper methysticum rhizoma (21 November 2017) — EU regulator.
- Assessment of the Risk of Hepatotoxicity with Kava Products. WHO, 2007 — WHO report produced with the Natural Standard Research Collaboration (per its press release); authors and funding not verified.
- FAO/WHO. Kava: a review of the safety of traditional and recreational beverage consumption, 2016 — prepared by a consultant (Biosearch Consulting) for FAO/WHO.
- FSANZ imported food risk statement: Kava (January 2025) — Australian/NZ government food regulator.
- Australia New Zealand Food Standards Code, Standard 2.6.3 Kava — legislation.
- Therapeutic Goods (Poisons Standard—October 2025) Instrument 2025 — Australian legislation.
- TGA listed-medicine compliance report: Noble Kava (Fiji Kava (Australia) Trading Pty Ltd), 2022 — Australian regulator.
- The Kava-kava in Food (England) Regulations 2002 (SI 2002/3169) — UK legislation.
- The Medicines for Human Use (Kava-kava) (Prohibition) Order 2002 (SI 2002/3170) — UK legislation.
- National Toxicology Program TR-571: kava kava extract (2012) — US government.
- Pittler MH, Ernst E. Kava extract versus placebo for treating anxiety. Cochrane CD003383 (full text) — UK; funding and declarations not re-checked this round.
- Sarris J et al. Kava for generalised anxiety disorder: a 16-week RCT. Aust N Z J Psychiatry 2020 — NHMRC APP1063383 with MediHerb/Integria co-funding (protocol, Trials 2015; NCT02219880; university news).
- Sarris J et al. Kava in the treatment of GAD (KALM). J Clin Psychopharmacol 2013 — registry funder Integria (ANZCTR).
- Sarris J et al. KALM adverse reactions, liver function, addiction and sexual effects. Phytother Res 2013 — same trial.
- Sarris J et al. KADSS. Psychopharmacology 2009 (publisher page) — registry funder University of Queensland (ANZCTR).
- Connor KM et al. Kava in GAD: three placebo-controlled trials. Int Clin Psychopharmacol 2006 — NCCIH grants (NCT00083980).
- Ooi SL et al. Kava for GAD: review. J Altern Complement Med 2018 — no funding statement in PubMed record.
- Volz HP, Kieser M. Kava-kava extract WS 1490 versus placebo: 25-week trial. Pharmacopsychiatry 1997 — affiliation not re-checked this round.
- Sarris J et al. WFSBP/CANMAT clinician guidelines. World J Biol Psychiatry 2022 — professional societies; lead author's industry payments declared.
- Lehrl S. WS 1490 in sleep disturbances with anxiety. J Affect Disord 2004 — funding not accessed.
- Sarris J et al. Acute effects of kava and oxazepam. Hum Psychopharmacol 2012 — funding not accessed.
- Russmann S et al. Hepatic injury due to traditional aqueous kava extracts in New Caledonia. Eur J Gastroenterol Hepatol 2003 — academic case series.
- Mathews JM et al. Inhibition of human cytochrome P450 by kava. Drug Metab Dispos 2002 — NIEHS contract.
- Gurley BJ et al. In vivo effects of goldenseal, kava, black cohosh and valerian on CYP enzymes. Clin Pharmacol Ther 2005 — NIGMS grant.
- Almeida JC, Grimsley EW. Interaction between kava and alprazolam. Ann Intern Med 1996 — case report.
- Foo H, Lemon J. Kava alone or with alcohol. Drug Alcohol Rev 1997 — academic.
- Norton SA, Ruze P. Kava dermopathy. J Am Acad Dermatol 1994 — academic.
- Sarris J et al. Does a medicinal dose of kava impair driving? Traffic Inj Prev 2013 — funding not accessed.
- Wainiqolo et al. Driving following kava use and road traffic injuries in Fiji. PLoS One 2016 — Wellcome Trust.
- Aporosa S et al. Traditional kava use and cognition: driver fitness. J Ethnopharmacol 2022 — academic.
- Economidis G et al. Global perspectives on kava. Drug Alcohol Rev 2025 — funder listed as the Australian Department of Health and Aged Care (Europe PMC record).
- Vanuatu Bureau of Statistics, Quarterly Statistical Indicators, December 2025 — government statistics.
- Fiji Times: Fiji's kava sector surges past $190M annually (February 2026) — newspaper citing official data.
- apotheke adhoc: BfArM allows kava only under strict conditions (September 2015); Pharmazeutische Zeitung (2015); NutraIngredients (2014) — German regulatory history (trade press).
- International Kava Organization press release (December 2025) — trade body; interested party.
- NutraIngredients: AHPA kava working group amid regulatory action (2025) — trade press.
- National Tribune: Integria Healthcare agreement (2020); FoodNavigator: Fiji Kava / The Calmer Co (2023) — company information.
- NCT05381025 — trial registration (Thorne HealthTech sponsor; results published 2024).
Educational research reviewed 6 October 2026. This article is not medical advice; talk to a qualified clinician or pharmacist before starting, stopping or combining treatments.
Have a question — or want us to cover something?
Ask about anything on this page, or request the next deep dive: an ingredient, a supplement, or a health concern. We use published research, evidence syntheses, and regulatory guidance, with clear source links.
One daily research roundup
Get the topics, key findings and links from our new articles in one email. At most one digest a day, only when there is something new.