Blood testing for THC comes up in a narrower set of situations than urine or oral fluid — mostly law enforcement, and mostly when the question is impairment, not general drug use. It's also the specimen type with the most scientific controversy attached to it, because the number a blood test produces doesn't mean what most people assume it means.
Why blood, and why here
Urine and hair are built for detecting drug use over days, weeks, or months. Blood is different: it reflects what's circulating in the body close to the moment of collection, which makes it the specimen of choice when the question is about the present moment — specifically, whether someone was impaired while driving, operating equipment, or at the time of an accident.
That's also why blood draws in this context usually happen under law enforcement authority rather than routine workplace screening. Drawing blood is more invasive than a cup or a swab, and it typically requires either consent or a warrant, which is part of why it's reserved for situations where the timing question genuinely matters — a traffic stop or a crash investigation, not a pre-employment check.
Why alcohol comparisons don't transfer
Blood alcohol concentration is the model most people have in their heads for what a blood drug test should do: a number goes in, an impairment level comes out, and the two move together in a predictable, well-studied way. The National Highway Traffic Safety Administration has documented exactly why that model works for alcohol and does not work for THC.
For alcohol, "the effects of alcohol consumption on behavior, judgement, cognition and emotions all correlate fairly well with the rise and fall of alcohol concentration in the body as measured by blood alcohol concentration and breath alcohol concentration," a relationship established through decades of controlled dosing studies where researchers measured impairment and blood alcohol together, repeatedly, across many people (NHTSA, Marijuana-Impaired Driving: A Report to Congress, DOT HS 812 440, July 2017). Higher blood alcohol reliably means more impairment. That relationship is why a single number — a blood alcohol content limit — can function as a legal standard on its own.
THC does not behave the same way. NHTSA's report is direct about the mismatch, describing the correlation of THC blood concentration with impairment as poor and noting that setting a per se limit the way states do for alcohol is not well supported by the underlying science. Part of the reason is pharmacokinetic: THC blood levels spike very high immediately after smoking and then fall rapidly, often dropping more than 80 percent within roughly the time it takes peak impairment to actually show up in performance testing. That means a person can test with a low THC blood level while still meaningfully impaired if enough time has passed since use, and a person can test with a high THC blood level from use that occurred recently enough that peak impairment hasn't caught up yet. High and low numbers alike can be poor stand-ins for what's actually happening in someone's coordination and judgment at that moment.
This is also why NHTSA notes that low THC concentrations — a few nanograms per milliliter — can reflect fairly recent use when impairment is plausible, while the same low number can just as easily reflect use from hours earlier when any impairment has already passed. The number alone, unlike blood alcohol, doesn't disambiguate between those two very different situations. Because that underlying relationship between concentration and impairment hasn't been established, NHTSA's own model drug-impaired-driving law does not include a THC per se provision (NHTSA, DOT HS 812 440, cited above; see also NHTSA, Drug-Impaired Driving — Understanding the Problem).
Per se laws exist anyway
Despite that science, a number of states have adopted per se or zero-tolerance THC laws for driving — meaning it's illegal to drive with THC in your blood above a specific threshold (or, in zero-tolerance states, any detectable amount at all), regardless of whether an officer observed signs of impairment. States that have set a specific numeric threshold include limits as low as 1 to 2 nanograms per milliliter and as high as 5, and several other states use zero-tolerance standards that don't rely on a threshold at all (NHTSA, Drug-Impaired Driving; NHTSA, DOT HS 812 440). Colorado instead uses a "permissible inference" law, where a THC concentration above a set level allows — but doesn't automatically require — a jury to infer impairment.
Why pass a law like that if the science behind the number is contested? Mostly for the same reason breathalyzer limits exist for alcohol at the practical, enforcement level: a threshold is easy to apply consistently, gives officers and prosecutors a clear standard to work from, and — per NHTSA's own research on officer experience — per se laws have been reported to have a measurable positive effect on prosecution and conviction outcomes in drug-impaired driving cases, even though officers didn't feel the laws made roadside enforcement itself any easier. It's a policy tool built around administrability, not a claim that a specific nanogram number is scientifically equivalent to a specific level of impairment.
This is genuinely contested legal terrain, and it's different in every state. If you're facing a specific legal situation involving a THC blood test — a DUI charge, a license question, anything with real consequences — that's a conversation for a lawyer licensed in your state, not something to sort out from a general explainer like this one.
Blood tests rarely stand alone
Because a THC number by itself is such a weak stand-in for impairment, law enforcement typically doesn't rely on a blood draw in isolation. Many agencies use officers trained through the Drug Evaluation and Classification Program, developed by the International Association of Chiefs of Police with NHTSA's support, who conduct a standardized, multi-step evaluation — including things like eye examinations, vital signs, and a structured interview — to form an opinion about impairment and the likely category of drug involved, before a toxicology result ever comes back (NHTSA, Drug Evaluation and Classification Program). The blood or oral fluid test result is meant to corroborate that trained observation, not substitute for it. That two-part structure — a documented behavioral evaluation plus a toxicology result — exists precisely because, as the research above shows, the number alone isn't a reliable enough signal on its own.
What a blood THC result does and doesn't tell you
A blood THC result tells you the concentration of THC (and sometimes its metabolites) circulating in blood at the moment of the draw. Because THC clears from blood relatively quickly after use compared with how long its metabolites linger in urine, a positive blood result generally does point to more recent use than a positive urine result would. But, per the research above, it is a poor stand-in for impairment on its own, in either direction. A number by itself — high or low — doesn't tell you how someone was actually functioning at that moment; that's a judgment the science genuinely struggles to make from blood concentration alone.
None of this is legal advice, and nothing here should be read as a way to predict or manage a test result. If you're looking at your own use and want a clearer sense of where you stand, our free quiz gives you a personalized usage tier and an estimated timeline in a couple of minutes.