Detection vs. Impairment — The Core Scientific Problem
Cannabis drug tests are almost universally designed to detect past exposure, not current impairment. This is the central scientific problem with cannabis drug testing: the tests we use and the questions we ask are not aligned.
The Simple Version
Alcohol breathalyzers measure current impairment. Cannabis urine tests measure whether you used cannabis days, weeks, or even months ago. These are fundamentally different measurements, but they are often used as if they were the same.
Why Alcohol and Cannabis Are Different
Alcohol is the archetype for "drug testing equals impairment testing." Breathalyzers detect the parent drug (ethanol) directly, and the concentration in breath tracks closely with the concentration in blood, which tracks closely with actual impairment. The relationship between blood alcohol concentration and impairment is well-established and roughly linear above threshold levels. A breathalyzer reading answers a meaningful question: is this person currently drunk?
Cannabis is different in almost every dimension:
- THC is highly lipophilic. It partitions out of blood into fat tissue rapidly, so blood THC concentrations drop sharply within hours of use even though the person is still impaired.
- The effects outlast the blood levels. Peak blood THC often occurs before peak subjective effect, especially with smoked cannabis.
- Metabolites persist far longer than effects. THC-COOH can remain detectable for days to weeks after effects are gone.
- Tolerance substantially affects impairment. Chronic users can have high blood THC concentrations with minimal functional impairment; occasional users can be meaningfully impaired at lower concentrations.
- Route of administration matters. Edibles produce different blood and metabolite profiles than smoking, even at the same total dose.
The result: no single blood, urine, saliva, or hair test reliably answers the question "is this person impaired right now?"
What the Tests Actually Measure
| Test Type | What It Detects | Relationship to Impairment |
|---|---|---|
| Urine | THC-COOH (inactive metabolite) | None. Can be positive weeks after last use. |
| Blood | Parent THC, 11-OH-THC, THC-COOH | Partial, short window. Best for recent use; unreliable in chronic users. |
| Saliva / Oral Fluid | Parent Δ9-THC | Partial. Correlates with recent use better than urine. |
| Hair | THC and THC-COOH in hair shaft | None. Reflects use over past 90+ days. |
| Sweat patch | Parent THC and metabolites | Continuous exposure window. Reflects any use during wear period. |
| Breath (experimental) | Δ9-THC on exhaled breath | Recent use only. ~2–3 hour window. Not yet approved for law enforcement. |
Only breath tests plausibly measure recent use in a way analogous to alcohol breathalyzers — and as of 2026, cannabis breathalyzers are not approved for law enforcement use and have limited commercial deployment. See Cannabis Breathalyzers.
The Consequences
For Workers
A worker who used cannabis on a Saturday night can be terminated on a Wednesday for a positive urine test, with no scientific basis for claiming they were impaired at work. Some states now specifically prohibit this:
- California AB 2188 prohibits discrimination based on tests detecting only nonpsychoactive metabolites
- New Jersey CREAMMA requires both a positive test and a physical evaluation by a certified Workplace Impairment Recognition Expert
- New York MRTA restricts marijuana testing unless a statutory exception applies
See Off-Duty Protections for the full state-by-state breakdown.
For Drivers (DUI)
Several states have enacted "per se" THC blood limits, typically 5 ng/mL. These are modeled on alcohol per se laws but rest on much weaker scientific foundations. The AAA Foundation for Traffic Safety concluded that per se THC laws are "arbitrary and unsupported by science." NHTSA's large case-controlled study found that drivers who tested positive for THC had identical crash risk to those testing negative once controlling for demographics and alcohol.
A 2025 Clinical Chemistry study found that 5.3% of regular cannabis users exceeded the 5 ng/mL blood THC limit at baseline after 48+ hours of abstinence, with no difference in driving simulator performance compared to users below the limit. See DUI & Driving.
For Probationers
The 30–90 day detection window in chronic users creates real injustice. A person entering probation who had used cannabis regularly for years may still test positive for weeks — triggering violations even with genuine abstinence. Several states (Connecticut, New Mexico, New Jersey, Minnesota, Missouri, Maine, Rhode Island) have modified probation cannabis testing in response. See Probation & Parole.
For Chronic Users Generally
The single most unfair situation in cannabis drug testing is the chronic user who stopped days or weeks before a test. They are not impaired. They may be actively trying to abstain. But the test will still be positive, and the person evaluating the test will usually not understand the difference.
What "Better" Testing Would Look Like
A scientifically defensible impairment test for cannabis would need to:
- Measure a property that correlates with current impairment, not past exposure
- Work across different routes of administration (smoking, edibles, vaping)
- Adjust for tolerance (the chronic user problem)
- Be practical to deploy in workplaces, roadside stops, and healthcare settings
Candidates under research include performance-based cognitive/psychomotor testing, cannabis breathalyzers, and combined biomarker approaches. None of these are yet mature enough for routine deployment. The current testing landscape is dominated by metabolite detection because it is logistically convenient, not because it is scientifically superior.