How Long After You Smoke Can You Get a DUI?

There is no number of hours that clears you. A cannabis DUI in the United States depends on what a test finds in your body and what your state's statute does with that finding, so the window of legal exposure runs from a few hours to roughly thirty days. Measurable driving impairment after smoking lasts about three to four hours. Blood THC in an occasional smoker usually falls below 5 ng/mL within a few hours, while a daily user can still test positive after a month of abstinence, and urine metabolites persist for weeks. In zero-tolerance states, any detectable trace is itself the offense.

What actually decides the answer, in six parts

The total is a sum, and each part carries its own number:

1. The statute written in the state you are driving through.
2. The specimen an officer can require: blood, oral fluid, or urine.
3. The delay between the stop and the collection of that specimen.
4. What you consumed, at what dose, and how recently.
5. How often you consume, which sets your baseline concentration.
6. Whether another way home existed at the moment you needed one.

Move any one of these and the answer moves with it. Someone who smokes twice a year in Sacramento and someone who medicates daily in Toledo ask the same question from two different legal realities.

Which statute applies where you are driving

California runs on impairment alone. Vehicle Code 23152(f) makes it unlawful "for a person who is under the influence of any drug to drive a vehicle," and the state has repeatedly declined to attach a number to it. No THC threshold convicts you in California, and none protects you. Prosecutors must prove you were actually affected, through driving pattern, officer observation, and field sobriety performance.

Washington runs on a number. Under RCW 46.61.502, you commit the offense if, within two hours after driving, your whole blood shows a THC concentration of 5.00 ng/mL or higher. The statute defines that concentration as delta-9-THC only and expressly excludes the metabolite THC-COOH. It also anticipates late draws. A sample taken beyond two hours may still be offered as evidence of your level inside the window, and any reading above 0.00 may be used as evidence you were affected by cannabis.

Colorado sits between the two. Under C.R.S. 42-4-1301, 5 ng/mL of delta-9-THC in whole blood creates a permissible inference of impairment. A jury may draw that conclusion and may equally refuse it; the prosecution still carries the burden of proving impairment.

I got this one wrong in print. Around 2015 I built a US road-trip comparison table with a column headed "legal limit" and put Colorado's 5 ng/mL in it beside state BAC figures. A reader in Denver flagged it, then a defense lawyer wrote in with the statutory language. A permissible inference is a rebuttable presumption, and that heading turned it into a hard rule for everyone who scanned the table instead of the note beneath it. The correction ran, the table came out, and the piece lost a syndication slot it had held for two years. Here the column heading does more damage than the number.

Illinois sets 5 ng/mL in whole blood or 10 ng/mL in another bodily substance, measured within two hours, under 625 ILCS 5/11-501. Ohio is the strictest structure in the country on this point. R.C. 4511.19 sets per se limits at 2 ng/mL of active THC in whole blood and 10 ng/mL in urine, and it separately criminalizes the inactive metabolite at 35 ng/mL in urine and 50 ng/mL in blood. An Ohio driver who consumed legally at home eleven days earlier can meet a per se element of the offense while stone-cold sober. Nevada moved the other direction in 2021: Assembly Bill 400 removed the 2 ng/mL per se thresholds from misdemeanor cases and kept them only where the charge is a felony.

The Governors Highway Safety Association, whose table was last updated in October 2025 with laws reviewed by state highway safety offices that June, counts four states with a specific per se THC limit, Colorado alone with a permissible inference, ten states with zero tolerance for THC or a metabolite, and four with zero tolerance for THC alone. Secondary compilations count differently, largely over how they handle Nevada's split statute. Check the statute for every state on your route, not a summary of it.

Blood, saliva, urine: what each specimen can prove

The three specimens answer three different questions, and readers routinely treat a result from one as if it settled the other two.

| Specimen | What it measures | Typical detection window | What a positive establishes | |---|---|---|---| | Whole blood | Delta-9-THC, the active compound | Hours in occasional users; up to 30 days in daily users (Bergamaschi et al., 2013) | The offense itself in per se states at or above threshold; elsewhere, evidence of use | | Oral fluid | Delta-9-THC deposited in the mouth | Roughly 8 to 24 hours occasional; up to about 72 hours frequent | A roadside screen, normally confirmed later by laboratory blood analysis | | Urine | THC-COOH, an inactive breakdown product | About 3 days occasional; several weeks daily | Past exposure, and in Ohio a per se element at 35 ng/mL |

Michigan's numbers show how the swab works in practice. Under Public Acts 242 and 243 of 2016, the Michigan State Police ran a one-year Oral Fluid Roadside Analysis Pilot in five counties, placing Alere DDS2 units with 31 Drug Recognition Experts from 12 agencies. Troopers ran 92 roadside tests, 89 drivers were arrested on observed behavior and the standardized twelve-step evaluation, and 83 tested positive on the swab. The swab supported the arrest; the twelve-step evaluation and the confirmatory blood draw did the legal work.

Here is what I cannot tell you. I have never watched a swab device run at an American roadside or sat through a Drug Recognition Expert evaluation, so my reading of these programs comes from agency reports and statutes. What I can vouch for, after eighteen years of comparing call-center scripts against published rules, is the failure mode that follows when a traveler assumes one jurisdiction's procedure resembles the one next door.

Why the hours between the stop and the blood draw matter

The specimen is collected long after the driving, and the number degrades in between. The AAA Foundation for Traffic Safety, analyzing Washington State Patrol data, found a median estimated time to blood draw of 165 minutes across all cases and 139 minutes among THC-positive drivers. Only 20 percent of THC-positive drivers reached 5 ng/mL at all. Split by delay, 26 percent of those drawn within two hours were at or above 5 ng/mL, against 10 percent drawn later.

That finding cuts against me, and the strongest version of it deserves stating before I answer it. Prosecutors and victim-advocacy organizations read those same numbers and conclude that per se thresholds are too lenient, because a driver genuinely impaired at the wheel routinely tests under the limit two and a half hours later. They are right about the mechanism, and the AAA data is theirs as much as mine. A blood result understates recent use at least as often as it overstates old use.

Which is the whole point. A measurement that errs in both directions, by an amount set by how quickly a phlebotomist was available, cannot be planned around by counting hours since the last session. Neither side of this argument produces a safe waiting period.

How long impairment lasts, and how long detection lasts

These two windows differ by more than two orders of magnitude, and the gap is where charges live.

For impairment, the reference point is the National Advanced Driving Simulator study by Hartman and colleagues, published in Drug and Alcohol Dependence in 2015. Occasional smokers drove 0.5 to 1.3 hours after inhaling vaporized cannabis. Blood THC of 8.2 and 13.1 µg/L during driving increased lane weaving by about as much as breath alcohol concentrations of 0.05 and 0.08. The measure was standard deviation of lateral position rather than driving performance in general, and cannabis with alcohol proved additive rather than synergistic.

For detection, the reference point is Bergamaschi and colleagues in Clinical Chemistry, 2013. Thirty male chronic daily smokers lived on a secure research unit for up to 33 days with daily blood collection. Twenty-seven of the thirty tested THC-positive on admission at a median of 1.4 µg/L. Two of the five participants still enrolled at day 30 remained THC-positive at 0.3 µg/L, and four of those five still carried detectable THC-COOH.

Four hours of impairment against thirty days of detectability is a ratio of roughly 180 to 1. NHTSA drew the obvious conclusion in its July 2017 Marijuana-Impaired Driving report to Congress, DOT HS 812 440, prepared under the FAST Act: blood THC correlates poorly with impairment, and setting per se levels on that basis is not meaningful.

I used to advise readers around a fixed waiting period. Until about 2019 the guidance I edited carried a six-hour rule, lifted from a state health department page that phrased it that way, and I let it stand because it was clean and easy to act on. I pulled it after reading the Washington statute properly. The offense there attaches to a concentration measured within two hours of driving, and says nothing about hours elapsed since your last session. A countdown clock answers a question no statute asks.

Edibles run on a different clock

Oral dosing inverts the relationship between how you feel and what you measure. Subjective effects begin 30 to 90 minutes after ingestion and reach maximum around two to three hours. Plasma concentrations lag and scatter. In Huestis's review of human cannabinoid pharmacokinetics, seventeen volunteers given a 10 mg Marinol capsule peaked at a mean of 3.8 ng/mL one to two hours in, while 15 to 20 mg of THC in sesame oil peaked around four to six hours. Double peaks are common, caused by enterohepatic recirculation, and the metabolites peak later still, at six to ten hours.

An edible can therefore leave you meaningfully affected while your blood THC sits at a level a smoker would reach on the way down. In a per se state that flatters you. In California, Colorado, or Nevada, where officer observations and roadside performance carry the case, it does not help at all.

The part you control: the ride you arranged

Five of the six parts are decided by other people. The sixth is yours, and the survey numbers say it is the one most often skipped.

The AAA Foundation surveyed 2,000 cannabis users across eight states and found 84.8 percent had driven within eight hours of consumption, with 53 percent driving within an hour or less. Colorado's Department of Transportation reported in 2025 that 33 percent of cannabis-using drivers had driven within two hours of use in the previous 30 days.

The alternative is thinner than most planning assumes. The American Public Transportation Association puts the share of Americans with no access to public transportation at 45 percent, and roughly 30 percent of rural areas have no transit service at all. A ride you intend to summon later is a hope about coverage in a place you have not checked.

What works is boring and mechanical:

1. Check the statute of every state on your route before you leave, including the ones you only pass through.
2. Arrange and confirm the ride before the first use, while the decision is still cheap.
3. If no ride exists at the destination, move consumption to the end of the travel day.
4. Stay out of the driver's seat entirely. Washington's RCW 46.61.504 and its equivalents elsewhere reach a person in physical control of a parked vehicle, keys in hand, engine off.

Eighteen years of editing disruption notices has taught me that the answer people need is rarely the number they asked for. The number here is unstable by design. The arrangement you make beforehand is not.

Questions people actually ask

How long after smoking can you get a DUI in California?

Indefinitely, in principle. California Vehicle Code 23152(f) prohibits driving under the influence of any drug and sets no numeric THC threshold, so no hour count protects you and none convicts you. Prosecutors must prove actual impairment at the time of driving, generally through driving pattern, officer observation, and field sobriety test performance.

Can you get a DUI with a medical card?

Yes. Washington's RCW 46.61.502 states directly that being entitled to use a drug under state law is no defense, and Ohio does not treat a medical marijuana card as a prescription for its per se provisions. Illinois is a partial exception, exempting registered patients from its 5 ng/mL presumption while leaving impairment charges fully available.

How long after eating edibles can you drive?

Longer than after smoking. Subjective effects start 30 to 90 minutes after ingestion and peak at two to three hours, while Huestis reports plasma THC peaking four to six hours after 15 to 20 mg of oral THC, with metabolites peaking later. Plan an edible as an evening without driving rather than an afternoon.

How much is five nanograms per milliliter?

Roughly five parts per billion, the same proportion as five seconds inside 31.7 years. Spread across the five liters of blood in an average adult, it amounts to about 25 micrograms of THC, well under one thousandth of the THC contained in a single half-gram joint at 20 percent potency.

Why does a positive test not answer impairment by itself?

Because THC persists far longer than its effects. NHTSA's 2017 report to Congress found blood THC correlates poorly with impairment, and Bergamaschi's 2013 study detected THC in daily users after 30 days of abstinence. A positive result establishes exposure. Impairment at the wheel is a separate question needing separate evidence.