Table of Contents
Cannabis driving laws come in three flavours: per-se limits that make a set blood THC level illegal by itself, zero-tolerance rules that ban any detectable THC, and impairment laws that ask whether you were actually too impaired to drive. Five states use per-se limits of 2–5 ng/mL, twelve use zero tolerance, and the rest rely on proof of impairment.
Laws change, and this is a map, not a legal opinion. Everything below reflects the rules as of November 2026, drawn from the National Conference of State Legislatures and federal driving research — check your own state’s statute before you rely on any of it. This article is general information, not legal advice.
The three ways states draw the line
Alcohol has a number everyone knows: 0.08% blood alcohol, and you are over the limit. Cannabis has no equivalent, because THC does not behave like alcohol in the body — and states have answered that problem in three very different ways.
A per-se limit sets a number in law. Cross it, and you are guilty whether or not your driving looked bad. A zero-tolerance rule goes further: any detectable THC (and in some states, its metabolites) in your blood is enough. An impairment law sets no number at all — the state has to prove, through your driving, a roadside test, and an officer’s observations, that you were actually too impaired to drive safely.
Every state prohibits driving while impaired by cannabis. The argument is not about that. It is about what counts as proof — and the three systems can treat the same driver, with the same blood test, in completely different ways.
Per-se states: the 2–5 ng/mL club
According to the National Conference of State Legislatures (NCSL), five states have per-se limits for THC in blood: Illinois, Montana, Nevada, Ohio, and Washington, with limits ranging from 2 to 5 nanograms per millilitre (ng/mL). Washington, Montana, and Illinois use 5 ng/mL; Ohio uses 2 ng/mL. Nevada’s 2 ng/mL figure applies to felony DUI cases rather than ordinary ones.
Colorado sits in a category of its own: a permissible-inference law. At 5 ng/mL or more, a jury is allowed to infer you were under the influence — but that inference can be challenged, and it is not an automatic conviction. Washington’s 5 ng/mL limit dates back to its 2012 legalisation initiative, making it one of the oldest cannabis driving numbers in the country.
Two details matter here. First, these limits measure THC in whole blood, usually drawn hours after the traffic stop — by which time the number has already fallen. Second, regular users can carry detectable THC for days after their last use, long after any high has faded — our guide to whether CBD shows up on a drug test explains how long THC traces can linger. A per-se law can therefore convict someone who was not impaired when they drove, which is exactly why researchers keep arguing about it.
Zero-tolerance states: any detectable THC
Twelve states take the hardest line, again per NCSL: Arizona, Delaware, Georgia, Indiana, Iowa, Michigan, Oklahoma, Pennsylvania, Rhode Island, South Dakota, Utah, and Wisconsin have zero-tolerance laws for certain drugs, including THC. In these states, any detectable amount can trigger a DUI charge — there is no number to stay under, because the number is zero.
Some of these states soften the rule for medical patients. Michigan, for example, amended its law in 2016 so that registered medical marijuana patients face an impairment standard instead of zero tolerance. Others offer no such exception at all. Arizona’s zero-tolerance rule dates to 2006, making it the first state to adopt one for cannabis.
Impairment states: prove it, don’t just measure it
The remaining states — including California, New York, Oregon, Massachusetts, and New Jersey — prosecute cannabis driving the way they prosecute any drug-impaired driving: on evidence of impairment. An officer’s observations, field sobriety tests, dashcam footage, and expert testimony all matter; a blood THC number on its own is not enough.
This sounds fairer, and in many ways it is — but it is also fuzzier. Field sobriety tests were designed for alcohol, not cannabis, and there is no roadside breath test for THC that works like a breathalyser. Officers lean on drug-recognition experts and blood draws, and cases turn into arguments between experts about what a given blood level actually meant at the time of driving.
That fuzziness is why the per-se debate never ends. Impairment laws risk letting impaired drivers go free; number-based laws risk convicting sober ones. No state has found a version that satisfies both sides.
Why a THC number is a shaky measure of impairment
The science problem is simple to state and hard to solve. After smoking, blood THC spikes, then falls fast — but impairment does not fall as fast. Research from the National Institute on Drug Abuse found that blood THC levels can drop below commonly used testing thresholds within a few hours, even while driving ability is still affected. A blood sample taken after the stop can therefore understate how impaired a driver was behind the wheel.
It cuts the other way too. Frequent users build tolerance and can carry blood THC above 5 ng/mL while driving normally, while an occasional user can be badly impaired below it — and THC can raise anxiety rather than calm it, as our guide to why weed makes some people anxious explains. A 2015 simulator study by Hartman and colleagues, funded by NIDA, found that blood THC of 8.2 and 13.1 ng/mL increased lane-weaving about as much as blood alcohol of 0.05 and 0.08 — but those were averages across 18 people, not a line any individual can rely on. The NIH’s own summary of the research puts it bluntly: THC concentration in blood is an imperfect measure of impairment while driving.
Edibles make the numbers look even stranger. A 2026 randomised trial in JAMA Network Open found that 10 mg and 20 mg THC edibles impaired simulator driving for at least five hours — while peak blood THC stayed at just 1.6 and 3.4 ng/mL, below every per-se limit in the country. The drivers were impaired; the number said they were fine. If you eat an edible and then drive, the law’s number may be the least of your problems.
What actually gets people caught
In practice, most cannabis DUI cases do not start with a blood test. They start with driving: weaving, speeding, a crash, a broken taillight — the same things that get any driver pulled over. Then the officer notices the smell, bloodshot eyes, or slow responses, and the stop becomes a DUI investigation.
Mixing cannabis with alcohol makes everything worse, and the research is clear on this point: the two together impair driving more than either one alone, and the effects add up even when each substance is below its own limit. If you have been drinking, the cannabis question is already answered — do not drive.
And remember the boring, practical rule that underlies all of this: impairment is not a feeling you can reliably judge in yourself. Cannabis can make people feel more cautious while actually slowing their reactions. If you feel any effect at all, do not drive — and if you took an edible, give it far longer than you think it needs. Our guide to driving after taking CBD covers the CBD side, where the rules are different but the caution is the same.
Medical card? It does not change the driving law
A medical marijuana card proves you can possess cannabis legally. It does not prove you were fit to drive, and in several states it is no defence at all. In Washington, for example, being a legal medical patient is no defence against the per-se limit. Some states do carve out exceptions for registered patients — Michigan’s zero-tolerance exemption is one — but you cannot assume your state is one of them.
This surprises people, because it feels backwards: the state said you could use it, so surely the state cannot punish you for driving after using it. But the driving law and the medical program are separate systems, written by different legislatures at different times, and they do not talk to each other. Check both before you drive.
Frequently asked questions
What is the legal THC limit for driving?
There is no national limit. Five states — Illinois, Montana, Nevada, Ohio and Washington — set per-se limits of 2–5 nanograms of THC per millilitre of blood, and Colorado lets a jury infer impairment at 5 ng/mL. Twelve states ban any detectable THC, and the rest require proof you were actually impaired.
Can you be charged with DUI if you were not impaired?
Yes, in per-se and zero-tolerance states. A blood test above the limit — or any detectable THC — can be enough on its own, even if your driving looked fine. THC can linger in blood long after the high fades, especially for frequent users, which is why these laws are so controversial.
Does a medical marijuana card protect you from a cannabis DUI?
Not reliably. Some states carve out exceptions for registered patients, but others do not — in Washington, for example, being a legal medical patient is no defence against the per-se limit. A card proves you can possess cannabis legally; it does not prove you were fit to drive.
How long after using cannabis is it safe to drive?
There is no fixed waiting time that works for everyone. Smoking impairs most in the first few hours, while edibles can impair driving for at least five hours after eating, even when blood THC is low. The honest rule is simple: if you feel any effect at all, do not drive — and never mix cannabis with alcohol.
Can CBD get you a DUI?
CBD itself is not intoxicating and has no per-se limit, so pure CBD should not trigger a cannabis DUI. The risk is THC hiding in full-spectrum products: a blood test measures THC, not CBD, and a zero-tolerance state will not care which bottle it came from. Read our guide to how to read a CBD lab report before you drive.
This article is general information, not legal advice. Cannabis driving laws vary by state and change frequently — check your state’s current statute and, if you are facing a charge, talk to a lawyer licensed in that state. See our medical disclaimer.
For a broader overview of legal and safety guides on this site, see our legal and safety hub.
Affiliate Disclosure
The CakeHeights Editorial Team is committed to providing accurate, research-backed information about cannabis and hemp products. Some links on this page are affiliate links, meaning we may earn a commission at no extra cost to you if you make a purchase through them. Our recommendations are based on independent research, testing, and analysis. We only promote products we believe offer genuine value to our readers. Learn more →