Most of what is written about kratom, by the people who sell it and by the people who want it banned, rests on evidence that cannot settle the question being argued. Pain relief, help with coming off opioids, low addiction risk, safe breathing: each of those claims leans mainly on rats, mice, cells in a dish, online surveys or individual case reports. We found no completed controlled trial of kratom in people who have the condition it is said to treat.
That gap runs in both directions. A claim of harm that rests on a rat study is weak for the same reason a claim of benefit that rests on a rat study is weak. This page applies one standard to both, and the honest answer to most of the questions below is that nobody knows yet.
This is an evidence review. It takes the claims a reader is most likely to meet and shows which tier of evidence each one stands on. It does not cover how the body absorbs and clears kratom’s alkaloids, which is a separate subject.
The tiers, and what each can show
The site grades evidence on five tiers, weakest first. The ones that matter here:
- In vitro (“in glass”) studies test a compound on cells or isolated receptors in a laboratory. They show what a molecule can do to a receptor. They say nothing about what happens in a body, where the compound has to be swallowed, absorbed, broken down and carried to the brain.
- Animal studies give a compound to rats, mice or fish. They show whether an effect exists in a living body, and at what dose. Doses are usually given per kilogram of body weight, often by injection rather than by mouth, and often as one purified alkaloid rather than the leaf. Together with in vitro work, this is tier 3 in our scheme.
- Surveys and case reports (tier 2) record what people say happened to them, or what doctors saw in one patient. A survey can show what people believe kratom does. It cannot show that kratom did it. A case report can show that something happened once. It cannot show how often.
- Human observational studies (tier 4) measure people who already use kratom, without assigning anyone to take it.
- Controlled trials (tier 5) give a measured dose of kratom, or a placebo, to people in a study and compare the two. This is the only design that can show kratom caused an effect in humans. For kratom it is rare, and the few that exist are small.
Why a rat result is not a human result
It would be easy to dismiss animal work entirely. That would also be wrong.
A 2024 umbrella review in PLoS Biology (Ineichen and colleagues), which drew on 122 systematic reviews covering 367 treatments across medicine, found that 50% of therapies tested in animals went on to any human study, 40% to a randomised trial, and 5% to regulatory approval, counting therapies at least ten years past their first animal experiment. For therapies studied in at least five animal studies, the authors report an 86% “concordance” between positive animal and positive human results, a pooled ratio of the two success rates rather than a count of matching therapies. We read the full text.
So animal results are not noise. They often point the right way. But only one in twenty of the therapies tested in animals was eventually approved for use in people, and that review covers medicine in general, not kratom. It tells you how far a rat finding usually is from a settled human answer. It does not tell you which kratom findings will hold.
There is also a reason specific to kratom. A 2019 study in mice (Kruegel and colleagues, ACS Central Science) found that mitragynine, kratom’s main alkaloid, is converted in liver preparations and in living mice into 7-hydroxymitragynine, a far more potent opioid-receptor agonist, and that brain levels of this metabolite were enough to explain most or all of mitragynine’s opioid-receptor-mediated pain relief in those mice. The authors write that this clarifies “the importance of route of administration”. Several of the animal studies below inject a purified alkaloid into a vein or the abdomen. People swallow a leaf that contains dozens of them (see kratom’s alkaloids).
The claims, one by one
“Kratom relieves pain”
Tier 3, in vitro and animal. A 2016 receptor study (Kruegel and colleagues, Journal of the American Chemical Society) found that mitragynine and 7-hydroxymitragynine are partial agonists at the human mu-opioid receptor, the receptor morphine acts on. Pain relief in rodents is widely reported: 110 of the records in our search combine kratom, pain terms and animal terms in their title or abstract. The 2019 mouse study above is one.
Tier 2, surveys. In a large US online survey (Garcia-Romeu and colleagues, 2020, Drug and Alcohol Dependence), 91% of 2,798 respondents said they used kratom for pain, and gave it high ratings of effectiveness. We read the full text. The authors call their sample self-selected: 53% heard of the survey through the American Kratom Association, an advocacy group, and 44% through social media. They say it is likely to under-represent people who stopped using kratom after a bad experience.
Tier 5, controlled trials. Two small studies, and they disagree.
- Vicknasingam and colleagues, 2020 (Yale Journal of Biology and Medicine) studied 26 men in Malaysia who had drunk kratom daily for six years on average. Each drank three drinks over one day in a random, double-blind order: 20 had one kratom drink and two placebos, and 6 had three placebos. Pain was induced with a cold-water hand bath. Pain tolerance rose from 11.2 to 24.9 seconds an hour after the kratom drink and did not rise after placebo. But the paper also reports that the difference between kratom and placebo at one hour was not statistically significant (p=0.12). The participants had stopped their usual kratom for 10 to 20 hours before dosing, and people with chronic pain were excluded. The authors say the result should be interpreted cautiously and does not endorse kratom for pain. We read the full text.
- Prevete and colleagues, 2025 (Psychopharmacology) gave synthetic mitragynine (5 to 40 mg) or placebo, single-blind, to 15 healthy volunteers in the Netherlands, and used the same cold-water test. Mitragynine had no significant effect on pain at any dose. The doses may be lower than regular users take, which the authors note. We read the full text.
A US government systematic review of plant-based treatments for chronic pain (the Agency for Healthcare Research and Quality, 2025 update, searching to April 2025) reports that no study it included evaluated kratom. We read this at abstract level; the full report would not load.
ClinicalTrials.gov lists one placebo-controlled trial in patients, of a mitragynine skin patch for chronic neck muscle pain, at a Thai university. On 29 October 2026 it was not yet recruiting. A completed, industry-sponsored trial of mitragynine products for muscle soreness after exercise in 100 healthy adults has posted no results.
What this means. Kratom’s alkaloids act on the receptor that opioid painkillers act on, and they relieve pain in rodents. Whether kratom relieves pain in people with pain has not yet been tested in a controlled trial with published results that we could find. It is not shown to work, and it is not shown not to work.
“Kratom helps people come off opioids”
Tier 3, animal. In a 2020 rat study (Hassan and colleagues, Frontiers in Psychiatry), rats were made dependent on morphine and then withdrawn. Mitragynine injected into the abdomen at 5 to 30 mg per kilogram reduced their withdrawal signs, as methadone and buprenorphine did at lower doses. There were six rats per group. We read the full text.
Tier 2, surveys. In the Garcia-Romeu survey, 1,144 respondents (41%) said they had used kratom to reduce or stop opioid use, and 411 attributed more than a year without opioids to it. This is the most quoted figure in the field, and it is a count of self-reports from a self-selected online sample.
Tier 5, controlled trials. None found. On 29 October 2026 ClinicalTrials.gov, the US trial registry, returned 15 records for kratom, mitragynine, Mitragyna speciosa or 7-hydroxymitragynine. None was a trial giving kratom or mitragynine to people as a treatment for opioid withdrawal or opioid use disorder. One recruiting study enrols people with past opioid use, but to compare kratom’s effects with oxycodone’s, not to treat anyone.
What this means. The idea rests on animals and on what users report. That is enough to justify a trial. It is not evidence that kratom works for opioid withdrawal, or that it is safer than the medicines that have been tested for it. Our withdrawal review covers the reverse question, withdrawal from kratom itself.
“Kratom is addictive”, and “kratom is not addictive”
Here the animal evidence argues with itself, which is a lesson in its own right.
Tier 3, animal, pointing towards low risk. In a 2018 study (Yue and colleagues, Psychopharmacology), rats trained to press a lever for methamphetamine would not keep pressing for mitragynine at any dose, which is one standard test of whether a drug is rewarding. A 2019 study (Hemby and colleagues, Addiction Biology) found that mitragynine neither replaced morphine for rats trained to self-administer it nor sustained self-administration on its own. We read only the abstracts of these two papers.
Tier 3, animal, pointing the other way. The same 2019 study found that rats readily self-administered 7-hydroxymitragynine, and its authors describe it as having “high abuse potential”. A 2021 study (Japarin and colleagues, Behavioural Brain Research, abstract only) found that mitragynine produced a conditioned place preference in rats, meaning they returned to the place where they had received it, and that mitragynine and morphine could each revive a preference the other had created. And the same group that showed mitragynine easing morphine withdrawal (Hassan and colleagues, 2021, Frontiers in Pharmacology) found that rats injected with mitragynine daily for 14 days showed withdrawal signs of their own when it was stopped, lasting at least five days. We read the full text.
Tier 5, controlled trials. Two, both small.
- Coe and colleagues, 2026 (Drug Testing and Analysis) gave dried kratom leaf in capsules or a placebo, double-blind, to healthy adults who had not used kratom for at least a year, once and then daily for 15 days. The paper analyses 116 people: 49 on kratom across four dose levels and 67 on placebo. It explains that the placebo group is larger because this was one formulation in a bigger study, which tested other formulations not reported, and the placebo groups were pooled. Ratings of drug effects rose with dose but differed significantly from placebo only at the highest single dose, and fell with repeated dosing. No participant reached the threshold for clinically meaningful opioid withdrawal after stopping. We read the full text. This is the same trial as the liver safety paper by Huestis and colleagues cited below, not a second study. The product’s maker paid for it, and every author discloses consulting for that company. Three consult through a firm that has also advised the American Kratom Association, and two disclose being paid expert witnesses in kratom litigation.
- Reissig and colleagues, 2026 (Journal of Clinical Psychopharmacology) gave single doses of 1 to 12 g of leaf kratom or placebo, double-blind, to 40 recreational polydrug users with opioid experience. The 12 g dose raised ratings of “drug liking”, “good effects” and “high”, which the authors describe as effects “commonly associated with drugs of abuse”. We read only the abstract.
Tier 2, surveys. In the Garcia-Romeu survey, 56 respondents (2%) met the diagnostic criteria for a moderate or severe kratom use disorder in the past year. A sample recruited largely through an advocacy group is not where you would expect to find the people most harmed.
What this means. In rats, mitragynine on its own looks weakly rewarding at most, while 7-hydroxymitragynine looks like a conventional opioid. In people, two small, short trials found abuse-related effects mainly at their highest doses. No study has followed people who use kratom for years to measure how many become dependent. Neither “addictive” nor “not addictive” is established.
“Kratom does not depress breathing like opioids do”
This is the claim most often offered as a reason kratom is safer than opioids, because the thing that kills people in an opioid overdose is breathing that slows and stops.
Tier 3, in vitro. The 2016 receptor study found that mitragynine and 7-hydroxymitragynine activate the mu-opioid receptor without recruiting a second signalling protein, beta-arrestin. That property has been proposed as a reason an opioid might cause less respiratory depression. It is a mechanism measured on receptors in cells, and it is a hypothesis about breathing, not a measurement of it.
Tier 3, animal. Three studies, and they do not agree about mitragynine.
- In rats, by mouth (Henningfield and colleagues, 2022, Psychopharmacology), mitragynine at 20 to 400 mg per kilogram produced no significant respiratory depression on blood-gas measures, while oxycodone did, with one death each at its two highest doses. The study was funded by the American Kratom Foundation, which the paper describes as established by the American Kratom Association, and its lead author’s firm consults for both. The authors write that their result “does not rule out” respiratory depression in some human conditions or in combination with other drugs. We read the full text.
- In mice, by mouth (Hill and colleagues, 2022, British Journal of Pharmacology), mitragynine did depress breathing at 10 mg per kilogram and above, but higher doses produced no further depression, a ceiling the authors attribute to the limited rate at which mitragynine is converted to 7-hydroxymitragynine. 7-hydroxymitragynine depressed breathing more with every dose increase. We read the full text.
- In rats, by injection into a vein (Zuarth Gonzalez and colleagues, 2025, Journal of Pharmacology and Experimental Therapeutics), 7-hydroxymitragynine depressed breathing and was 4.5 times more potent than morphine at halving the volume of air breathed per minute, while mitragynine increased breathing rate. Naloxone reversed the depression. We read only the abstract.
The animal studies do agree on one thing: 7-hydroxymitragynine depresses breathing in a dose-dependent way, as conventional opioids do. Our page on concentrated 7-OH products explains why that matters.
Tier 4 and 5, human. No human study has been set up to test whether kratom depresses breathing. Some trials recorded breathing rate as a routine safety check. In the Prevete trial, 20 mg of mitragynine slightly reduced breathing rate compared with placebo, within the normal range. In a 2024 study at the US National Institute on Drug Abuse (Smith and colleagues, Journal of Addiction Medicine), ten regular users took their usual dose in the clinic, without blinding, and their oxygen saturation and breathing rate did not change significantly. We read both in full.
Tier 2, case reports. A 2026 systematic review of kratom and the lungs (Kanchustambham and Saladi, Journal of Medical Toxicology) found six published cases of opioid-type respiratory depression, all of whom survived. The reviewers graded three as probably caused by kratom, including one with high urine levels of 7-hydroxymitragynine, and three as possibly caused. Each of those three involved other substances, and one was a concentrated 7-hydroxymitragynine product rather than leaf. We read the full text.
What this means. The reassurance rests on a receptor mechanism and on animal studies that disagree with each other about mitragynine. The human evidence is a handful of breathing-rate readings in small trials that were not looking for respiratory depression, and six case reports. Whether kratom depresses breathing in people at the doses they take, or with other drugs, has not been measured.
“Kratom damages the liver”
This claim is the exception: its evidence is mostly human, not animal. Our liver review covers it in full.
The strongest single source is a US research network that investigates suspected drug-induced liver injury. In 2020 it reported 11 cases attributed to kratom after formal causality assessment, all of whom recovered (Ahmad and colleagues, Drug and Alcohol Dependence). That is tier 2: it shows kratom can injure the liver, and nothing about how often. In the industry-funded leaf trial above (Huestis and colleagues, 2026, Therapeutic Drug Monitoring), raised levels of the liver enzyme ALT were among the most common adverse events during daily dosing, affecting 3 of 12 participants at the highest dose against 1 of 60 on placebo. We read both in full. The liver review also covers the one rat study we found.
“Kratom lifts mood and anxiety”
In the Garcia-Romeu survey, 67% said they used kratom for anxiety and 65% for depression. That is tier 2. In the Prevete trial, the lowest dose of mitragynine (5 mg) raised self-rated arousal and attention in healthy volunteers, and the highest (40 mg) raised ratings of amnesia and produced mild symptoms of distress. We found no controlled trial in people with anxiety or depression.
“Kratom kills”
Death statistics are observational data, and they record whether mitragynine was found or named, not whether it caused the death. That distinction, and what the coroner reports show, is the subject of our mortality review. In the 2026 lung review above, 15 of the 24 reports, fatal and non-fatal, involved other substances as well as kratom.
The claims, side by side
| Claim | Strongest evidence we found | Tier | What it cannot tell you |
|---|---|---|---|
| Relieves pain | Two small controlled studies in volunteers, which disagree | 5, but not in patients | Whether it relieves pain in people who have pain |
| Helps people come off opioids | One rat study; one large self-selected survey | 3 and 2 | Whether it works, or how it compares with tested medicines |
| Low addiction risk | Rat self-administration studies; two small, short trials | 3 and 5 | What happens over years of use |
| High addiction risk | Rat studies of 7-OH and place preference; survey rates of use disorder | 3 and 2 | How many users become dependent |
| Does not depress breathing | Receptor assays; animal studies that disagree; routine breathing-rate readings in small trials | 3, some 5 | What happens at real doses, or with other drugs |
| Depresses breathing | Animal studies of 7-OH; six case reports | 3 and 2 | How often, and at what dose, in people |
| Injures the liver | Registry case series; enzyme rises in one trial | 2 and 5 | How often |
| Lifts mood or anxiety | Surveys; one small volunteer trial | 2 and 5, not in patients | Whether it treats either condition |
The gap cuts both ways
It is tempting to read a table like this selectively. Someone who wants kratom left alone will notice that the breathing and addiction harms rest largely on rats. Someone who wants it banned will notice that the pain and opioid-withdrawal benefits rest on rats and on surveys of enthusiasts. Both observations are correct. Thin evidence of harm is not evidence of safety, and thin evidence of benefit is not proof of benefit. An untested claim does not become true because its opposite is also untested.
The gap is not the same everywhere, either. Some of the harm evidence is human: the liver registry cases, the enzyme rises in a placebo-controlled trial, six case reports of slowed breathing, and abuse-related effects at the top doses of two small trials. None of that tells you how common a harm is. On the benefit side we found nothing equivalent in patients, only volunteers and surveys. Where the human evidence leans one way, this page says so rather than splitting the difference.
Two more things are worth carrying into anything else you read about kratom.
Who paid. Several of the studies above that point towards safety were funded by, or written by consultants to, a kratom product maker or a kratom advocacy group: the leaf capsule trial and the oral rat breathing study among them. That does not make their data wrong, and their funding is disclosed in the papers. It is a reason to want independent replication. Some of the work was publicly or academically funded, including the Malaysian pain study, the US direct-observation study (one of whose authors discloses consulting for the leaf trial’s sponsor) and the large survey’s authors’ grants. A conflict of interest can point either way, which is why we publish our own on every page.
Counting papers is not counting studies. The leaf capsule trial has produced at least three papers in our search: one on liver safety, one on abuse potential and one on how the body handles the alkaloids. A reader who counted those as three trials would triple the controlled evidence.
How we found these papers
Europe PMC, a European database of biomedical literature, was searched on 29 October 2026. The base
query TITLE_ABS:(kratom OR mitragynine OR "Mitragyna speciosa" OR "7-hydroxymitragynine")
returned 1,189 records. Adding claim terms to the base query returned:
- 391 for
(pain OR analgesia OR analgesic OR antinociception OR antinociceptive OR nociception OR nociceptive) - 210 for
("opioid withdrawal" OR "opioid use disorder" OR "opiate withdrawal" OR "morphine withdrawal" OR "opioid dependence" OR "opioid addiction") - 203 for
("abuse liability" OR "abuse potential" OR "self-administration" OR "conditioned place preference" OR reinforcing OR addiction OR addictive) - 52 for
(respiratory OR respiration OR breathing OR hypoventilation)
Adding the animal terms TITLE_ABS:(rat OR rats OR mice OR mouse OR murine OR rodent OR rodents OR zebrafish OR monkeys OR animal OR animals) to the pain search returned 110.
To find controlled human studies, the base query was combined with TITLE_ABS:(randomized OR randomised OR placebo OR "double-blind" OR "healthy volunteers"), which returned 27 records, and
separately with terms for crossover, phase 1 and clinical-trial publication types, which returned 21
more. We read every title. Eight of the 27 report giving kratom, mitragynine or a kratom blend to
people in a study. Three of those eight are papers from the single leaf capsule trial, so they
describe six studies, five of them with a placebo. None enrolled patients with the condition kratom
is said to treat. Further animal studies were found with searches for self-administration, place
preference, withdrawal and respiratory terms. ClinicalTrials.gov was searched on the same day for
kratom OR mitragynine OR "Mitragyna speciosa" OR 7-hydroxymitragynine and returned 15 records,
of which the neck-pain patch trial was the only one in patients with a condition kratom is said to
treat.
A keyword search can miss a study whose abstract uses other words, and we did not search for non-English papers. If a controlled trial in patients exists and we missed it, the conclusions about pain and opioid withdrawal would change, and we would correct this page.
Reading depth. We read in full: the translation review, both Kruegel papers, the two pain studies, the survey, both Hassan rat studies, the leaf capsule trial’s liver and abuse-potential papers, the oral rat and mouse breathing studies, the direct-observation study, the lung review and the liver registry series. We read only the abstracts of the Yue, Hemby, Japarin and Reissig papers, the intravenous rat breathing study and the 2025 AHRQ review, and the claims drawn from them go no further than their abstracts. The Japarin paper carries a 2023 corrigendum, which we read: it adds post-conditioning data that had been left out of the figures, and does not change the findings described here.
This page is not medical advice, and nothing on it is a reason to use or to avoid kratom. It describes what has and has not been measured.