In Wisconsin, driving with any detectable amount of mitragynine in your blood is an offence, and the prosecution does not have to prove that you were impaired. In England and Wales the opposite is true: there is no number, and a court has to be persuaded that your ability to drive was actually impaired.
Those are the two shapes drug-driving law comes in, and which one a reader is standing in front of matters more than anything the research says. It matters more because the research is thin. The only measurement of kratom and driving performance in the published literature is a pilot study of ten people on a simulator, and it found no significant change.
How this review was put together
We searched Europe PMC on 2 September 2026 for records combining kratom or mitragynine with driving, psychomotor performance, impaired driving, DUID or motor vehicle. The search returned 461 records.
That search string had a hole in it, and we record it because the hole is instructive: it did not contain the words cognitive, cognition or neurocognitive, even though one of the papers it did return is titled for its neurocognitive outcomes. Re-running it with those terms surfaced a study with a non-user control group, which is the study design the rest of this literature lacks.
Nine records are used below. We read the full text of six of them. Only one, the Orange County casework, is cited at abstract level, because it is behind a paywall with no author manuscript; everything attributed to it here comes from its published abstract and is marked as such.
We did not find a randomised controlled trial of kratom and driving, and we do not think one exists.
We label each block by evidence tier, weakest first, because averaging a survey with a trial produces a number that means nothing.
What people say about their own driving (tier: cross-sectional self-report and momentary assessment)
Zamarripa and colleagues, publishing in Traffic Injury Prevention in 2024, ran a remote study of US adults who use kratom regularly. 395 completed the survey and 357 completed a 15-day ecological momentary assessment, meaning repeated reports collected in daily life rather than recalled afterwards. Participants who did not finish the assessment, and those who said they do not drive, were excluded from the driving analyses, which leaves 309.
Of those 309, 284 reported having driven under the influence of kratom. That is 92%. Most reported low rates of risky driving and high confidence in their driving afterwards, and the momentary reports agreed: within 15 to 180 minutes of using kratom, participants reported feeling confident and perceived little impairment.
What this measures is belief. A driver’s confidence in their own ability is a poor instrument for detecting impairment, and it is the measure most likely to be wrong in the reassuring direction. The finding worth carrying forward is the exposure figure. Driving after kratom is routine among people who use it regularly.
One laboratory session, reported twice (tier: pilot study, n = 10)
Two of the papers in this literature look like independent corroboration and are not. They are the same ten people, in the same clinic, in the same sessions, split across two write-ups. Both carry the trial registration NCT05457803. Smith’s paper says its non-uniform timepoints were a consequence of “scheduling the driving simulator” and that outcomes “reported elsewhere included simulated-driving performance”; Zamarripa’s says findings unrelated to driving are reported elsewhere. Anyone counting them as two studies has doubled a sample of ten.
Here is what that one cohort produced.
On the simulator, nothing changed. Ten adults, re-recruited from the remote study, consumed their own typical kratom products and drove a high-fidelity simulator before and after. There were no significant changes in simulated driving performance.
That is the only measurement of kratom and driving performance we could find in the published literature, and it needs reading carefully rather than quoting. Ten people is a pilot. The participants were regular consumers, the group most likely to have developed tolerance, taking doses they chose themselves. The authors record that there was no placebo condition and no control group, and their own conclusion is that research is needed to characterise kratom’s effect on driving in regular and infrequent consumers. A study of this size can fail to detect a real effect. It cannot demonstrate the absence of one.
Off the simulator, two things did change. The mean dose was 5.16 grams of leaf powder, range 1.1 to 10.9 grams. Pupil diameter decreased between 40 and 80 minutes after dosing and stayed below baseline beyond 160 minutes, which is the classic opioid sign. And although the paper’s summary says psychomotor performance did not reliably improve or deteriorate, its results section reports that mean reaction time was significantly higher than baseline 80 to 120 minutes after dosing (b = 23.1, p = .01), though not at any other timepoint.
A slower reaction time in the second hour is the single most driving-relevant measurement in this cohort, and it does not appear in the abstract. We include it because a page about driving that reported only the summary sentence would be reporting the reassuring half.
What controlled dosing does (tier: phase 1 trial, non-driving outcomes)
Prevete and colleagues ran a placebo-controlled, single-blind, within-subjects phase 1 study published in Psychopharmacology. Eight healthy volunteers received placebo and 5, 10 and 20 mg of mitragynine on separate days, in a sequential rather than randomised order. A second group of seven received placebo and 40 mg. Measurements ran for seven hours.
Mitragynine did not affect most outcomes at any dose. The exceptions matter, and one of them cuts against the reassuring reading.
At 5 mg, attention improved and inhibition got worse. The paper reports that 5 mg significantly reduced the number of attentional lapses in the psychomotor vigilance task (p = .024) but increased the number of errors in the stop-signal task (p = .035). The discussion states it plainly: the low dose “improved accuracy in a sustained attention task and decreased inhibition in the stop signal task”.
Failing to withhold a prepared response is what a stop-signal error is. It is also, in outline, what braking for a hazard requires. This is the closest thing to a driving-relevant impairment signal anywhere in the human literature, and it appeared at the lowest dose tested.
At 40 mg, subjective ratings of amnesia increased and mild psychopathological symptoms appeared. Doses up to 40 mg were well tolerated and side effects were mild and transient.
The trial also measured blood levels, and they are worth putting next to the roadside numbers below. Peak plasma mitragynine after 1 hour was about 5.7, 15.7 and 25.6 ng/mL at 5, 10 and 20 mg.
Two limits on all of this. It measured no driving. And it used isolated mitragynine, not kratom: a leaf product contains mitragynine alongside a set of other alkaloids, and the doses people take are not easy to map onto milligrams of one compound. See our page on the kratom alkaloids.
The one study with a control group (tier: cross-sectional, human, non-driving)
Singh and colleagues, in the Journal of Psychoactive Drugs in 2019, compared 70 regular kratom users with 25 control participants on the Cambridge Neuropsychological Test Automated Battery, across six tasks covering motor function, learning and memory, attention and executive function.
Higher consumption, more than three glasses of kratom tea daily or mitragynine doses between 72.5 and 74.9 mg, was selectively associated with impaired performance on the Paired Associates Learning task, which measures visual episodic memory and new learning. On every other domain, users and controls performed comparably, and so did heavy and light users.
This is long-term use rather than acute effect, and it is not a driving study. It is here because it is the only human kratom study we found with a group of people who do not use kratom to compare against, and its finding is narrow: one deficit, in memory, at the top of the dose range.
What roadside and crash toxicology finds (tier: forensic casework and prevalence surveys)
Arrested drivers. Kedzierski and Mata examined three years of casework from Orange County, California. Of 25,398 cases received between 2017 and 2019, 60, or 0.24%, had mitragynine above 10 ng/mL. Concentrations ran from 10.5 to 960 ng/mL, mean 109, median 58. Of the 44 cases also screened for 7-hydroxymitragynine, 27 were positive. Ninety per cent of the drivers were men. No case contained mitragynine alone. The drugs most often found alongside it were central nervous system depressants, led by ethanol and then benzodiazepines, then stimulants such as methamphetamine and cocaine, then opioids including fentanyl. The authors discuss the two cases containing only one other psychoactive substance, which indicates how rarely the data permit even that. This paragraph is taken from the published abstract; we could not obtain the full text.
Put the median next to the trial figures above. 58 ng/mL is more than twice the peak plasma level 20 mg of isolated mitragynine produced, and the top of the range is roughly forty times it. Whatever those drivers had taken, it was not a 20 mg dose.
Crash victims. Gerona and colleagues screened serum from 1,000 roadway crash victims in California against a database of 1,314 drugs including 1,008 new psychoactive substances. Eight new psychoactive substances were confirmed across 17 cases. Mitragynine was the third most frequently detected, after bromazolam and para-fluorofentanyl. Its serum concentrations had a geometric mean of 7.02 ng/mL and a range of 0.55 to 90.55, lower than levels reported in overdose and death cases.
One detail in that paper’s table is a genuine counterweight to the Orange County finding. Of its mitragynine cases, three co-detected other drugs, but one, a 23-year-old male driver at 0.55 ng/mL, was negative for everything else. “Never alone” is a finding about one county’s arrested drivers, not a law of nature.
A single case report. Wright, in 2018, published what it describes as the first reported suspected driving-under-the-influence case involving mitragynine: a 37-year-old woman who nearly struck an oncoming vehicle. The drug recognition expert concluded she was under the influence of a central nervous system stimulant and cannabis, and mitragynine was only found because she told the officer she had used kratom, prompting testing outside the normal protocol. Amphetamine was quantified at 0.052 mg/L, and mitragynine and citalopram were reported qualitatively. It is a case report with two confounders, and its value is procedural: the substance was found because the driver volunteered it.
Unselected drivers. Ingsathit and colleagues stopped 1,635 randomly selected drivers across five regions of Thailand in 2005 and 2006 and tested their urine. Psychoactive drugs were present in 9.7% of samples. Among illicit drugs: amphetamine 1.8%, cannabis 1.1%, mitragynine 0.9%, morphine 0.1%.
That 0.9% is the only figure here drawn from drivers who were not stopped on suspicion or involved in a crash, and it is several times the 0.24% found among arrested drivers in Orange County. The two are not directly comparable, and the reason is the point: a case series of arrested drivers is selected on the outcome, and cannot tell you whether a drug causes impairment.
Would a test even find it?
In the United States, usually not.
The National Safety Council’s Alcohol, Drugs and Impairment Division publishes the consensus recommendations that US and Canadian laboratories use to decide what to test for in impaired driving cases. The 2021 update sorts drugs into two tiers. Tier I compounds were “considered essential for inclusion in routine testing workflows”. Tier II analytes “had limited or regional prevalence”, and laboratories “may or may not elect to include” them depending on regional trends and laboratory resources.
Mitragynine is in Tier II. A laboratory following the recommendations may test a driver’s blood without ever looking for it.
The survey behind the tiers puts a number on how rare that testing is. Its table counts, for each drug, how many of 64 responding laboratories listed it among their own fifteen most frequently detected drugs. Fentanyl scored 45 and amphetamine 43. Mitragynine scored 4.
The practical consequence is that every prevalence figure on this page is a lower bound. A substance most laboratories do not screen for is a substance that mostly does not appear in the statistics.
One further detail: the recommendations place mitragynine under the drug recognition expert category CNS stimulants. So does Wisconsin’s statute, which lists it under the heading STIMULANTS. That sits oddly beside its opioid receptor pharmacology, and beside the Orange County finding that it turns up mostly with depressants.
The law asks two different questions
An impairment offence requires proof that the driver’s ability was impaired. It applies to whatever caused the impairment.
A per se offence requires proof that a named substance was present, sometimes above a stated concentration and sometimes at any detectable level. Impairment does not have to be proved. These offences exist only for substances a legislature has named.
Which one applies to kratom depends entirely on whether the jurisdiction has scheduled its alkaloids. Below are the two we read in full, one of each kind.
Wisconsin: a detectable amount is the offence
Three provisions have to be read together.
Section 961.14(7) of the Wisconsin Statutes is Schedule I, subsection (7), headed STIMULANTS. Paragraph (mk) is “Mitragynine” and paragraph (mL) is “7-hydroxymitragynine”. Neither entry carries a threshold.
Section 340.01(50m)(a) defines a “restricted controlled substance”, for the motor vehicle code, as “a controlled substance included in schedule I under ch. 961 other than a tetrahydrocannabinol”. Mitragynine is in schedule I and is not a tetrahydrocannabinol.
Section 346.63(1)(am) makes it an offence to drive or operate a motor vehicle while “the person has a detectable amount of a restricted controlled substance in his or her blood”.
There is no threshold anywhere in that chain, and impairment is not an element. The annotation printed with section 340.01 states the courts’ reading directly: subsection (50m) “is part of a statutory scheme that creates a strict liability, zero tolerance approach to driving a motor vehicle after illegally ingesting a restricted controlled substance, without regard to impairment”, citing State v. VanderGalien, 2024 WI App 4.
England and Wales: impairment, and only impairment
Section 4 of the Road Traffic Act 1988 makes it an offence to drive, attempt to drive or be in charge of a mechanically propelled vehicle on a road or other public place while unfit to drive through drink or drugs. Section 4(5) says a person is unfit if their ability to drive properly is for the time being impaired. Section 11(2) defines “drug” to include any intoxicant other than alcohol.
That definition is wide enough that nothing turns on kratom’s legal classification. If it impairs the driver, section 4 reaches it.
Section 5A is the per se offence, inserted by the Crime and Courts Act 2013. It applies only to a specified controlled drug above a specified limit, both set by the Drug Driving (Specified Limits) (England and Wales) Regulations 2014, in force from 2 March 2015. As made, that table listed sixteen drugs. Amphetamine was added on 14 April 2015 by S.I. 2015/911, making seventeen: amphetamine, benzoylecgonine, clonazepam, cocaine, delta-9-tetrahydrocannabinol, diazepam, flunitrazepam, ketamine, lorazepam, lysergic acid diethylamide, methadone, methylamphetamine, methylenedioxymethamphetamine, 6-monoacetylmorphine, morphine, oxazepam and temazepam.
Mitragynine is not among them, in either version.
Section 5A also requires the substance to be a controlled drug within the meaning of section 2 of the Misuse of Drugs Act 1971. Kratom is not controlled under that Act; Great Britain restricts it through the Psychoactive Substances Act 2016 instead, which our United Kingdom page sets out. So the per se route is closed twice over, and section 4 is the only one open.
Everywhere else, and what we did not check
Impaired driving in the United States is state law, and every state has an offence that turns on driving under the influence of a drug. Where a state also has a per se or zero-tolerance provision, it is generally keyed to controlled substances, so it reaches kratom only in states that have scheduled its alkaloids. Our United States section covers the federal position and the states we have read so far, which does not yet include Wisconsin.
Wisconsin is not the general case, and we are not claiming it is. We read its three provisions because a fact-check of an earlier draft of this page found them. We have not read the driving statutes of the other states that schedule mitragynine, and we have not surveyed European per se lists. An earlier version of this page said no jurisdiction anywhere sets a limit for kratom. That was wrong, and it was wrong in the direction that would have misled a reader in front of a police officer.
What we do not know
The gaps remain larger than the findings.
- No dose-response study of driving. The one simulator study used doses participants chose for themselves.
- No acute study of people who do not use kratom. Prevete’s volunteers were kratom-naive, but they were given isolated mitragynine and never drove anything. Singh’s controls did not take kratom at all.
- No driving or psychomotor data on concentrated products. Human pharmacokinetic work on kratom extract now exists; nothing pairs an extract or a 7-OH-enriched product with a performance measure.
- No data on kratom with alcohol or benzodiazepines, which is the combination roadside casework actually finds.
- No validated impairment threshold. The Orange County concentrations span two orders of magnitude and no evidence links any of them to a degree of impairment.
The honest verdict
The strongest statement the evidence supports is narrow, and narrower than it looked before the papers were read in full. One cohort of ten regular consumers, taking their usual doses, showed no measurable change in simulated driving performance, and slower reaction times in the second hour afterwards. In a phase 1 trial, the lowest dose of isolated mitragynine sharpened sustained attention and worsened response inhibition.
That is not kratom being shown not to impair driving. It is a very small amount of mixed evidence, none of which speaks to a first-time user, a concentrated extract, a large dose, or kratom taken with anything else.
The law does not wait for any of it. In England and Wales an impairment offence reaches any intoxicant, and the prosecution has to prove the effect. In Wisconsin a detectable trace is the whole offence. The absence of a per se limit in a given place is not a permission; it only means that nothing there can be proved by a number, in either direction.
If a study we have missed bears on this, or a jurisdiction adds mitragynine to a per se list, write to editorial@wikikratom.com and we will update the page and credit the correction.