Module · Body

Drug Test
Detection Times

Estimated detection windows across urine, blood and hair, averaged from ten federal and academic sources. How often you used is the input that moves the answer most, so it's weighted heaviest here. These are ranges rather than dates, because that's genuinely all the underlying data supports.

See the math behind this calculator

Matrices covered
Urine · Blood · Hair

Biggest single factor
Use frequency

What this is
An estimate

The math

There isn't a real equation here, and pretending otherwise would be the dishonest move. Detection windows come from published clinical ranges, not from a formula you can derive. What this page does is pick the right published range for your pattern of use and lay it against your last-use time.

window = published range for [substance][frequency]
adjusted = window × body-mass factor (cannabis only)
window ends = last use + upper bound

Your weight and height are asked for, and then deliberately not used to change anything. That needs explaining. THC really does store in fat, so it's reasonable to expect a heavier person to test positive for longer, and an earlier version of this page multiplied the cannabis windows by a BMI factor on exactly that reasoning. The number behind it was invented rather than measured, and when we went looking for something to justify it we found the opposite. Schuster and colleagues tracked 62 cannabis users through a month of verified abstinence and reported that sex, BMI, race, years of use and past-month frequency were all not significantly associated with how fast THC-COOH cleared. So the factor is gone, the windows are exactly what the sources say, and the readout shows your BMI with no adjustment attached to it.

Frequently asked

Why is the marijuana window so much longer than everything else?

Because THC is fat-soluble and almost nothing else on this page is. Cocaine, meth, heroin and opioid pills are cleared through water-based routes within days. THC parks itself in fatty tissue and leaches back out slowly, so the metabolite that urine tests look for keeps appearing long after the effects are gone. That's also why frequency matters so much more for cannabis - each session tops the store back up. A first-time user is usually clear in a few days, while someone smoking several times a day can still test positive a month later.

Does drinking water or exercising shorten the window?

Not meaningfully, and both can backfire. Drinking a lot of water dilutes your urine rather than removing anything, and labs test for that directly - a sample with low creatinine and low specific gravity is flagged as dilute, which is usually treated as a failed or invalid test rather than a pass. Exercise mobilises fat and can briefly raise the THC metabolite in your blood and urine, so training hard the day before a test can push a borderline result the wrong way. Nothing on the market reliably speeds elimination up.

How far back does a hair test really go?

Two of our sources give the same answer on the front of that window - use does not show until it has grown 7 to 10 days out of the scalp, so a hair test is close to blind to the past week and a half. On the back of it they spread: three put the window at roughly 90 days and SAMHSA TIP 47 puts it at four to six months, which averages to about 105 days. Hair is also poor at catching one-off use, since it detects patterns rather than single occasions. And it is worth knowing that hair is not an approved specimen for federally regulated workplace testing at all.

Will this tell me whether I'll pass a specific test?

No, and no calculator can. These are population ranges drawn from published detection windows, and individual variation is enormous - metabolism, kidney and liver function, dose and potency, how it was taken, hydration, body composition and other medications all move the number. The test itself matters too, because the cutoff differs between panels and a lab confirmation test is far more sensitive than the initial screen. Treat the output as a rough sense of scale rather than a date to plan around.

Show the working

This is a lookup of published figures with a date attached. Some bands are averaged straight from the sources and some are interpolated between them, and the list below says which is which. Everything here is what the page actually does. If any of it looks wrong, it might be - tell us what we got wrong.

The lookup

low  = mean of each source's lower bound
high = mean of each source's upper bound
(sources reporting only "up to X" feed the high, not the low)

Body mass — computed, but not applied

BMI = weight kg ÷ (height m)²
window = unchanged by BMI

Which bands are sourced

cannabis   first, daily, heavy   sourced
cocaine    typical, heavy          sourced
heroin     typical, heavy          sourced
meth       typical, heavy          sourced
alcohol    EtG, ethanol, blood     sourced
opioids    typical                 sourced
everything else                    interpolated between them

Turning a window into dates

earliest clear = last use + low
window ends    = last use + high

Hair, all substances window = 8.5 days … 105 days after use (source average)

Where the ranges come from

Ten published sources — nine US federal or NIH publications and one university public-health department, listed in full below
No body-mass coefficient — an earlier version multiplied the cannabis windows by BMI. That number was invented, and the evidence does not support it, so it is gone
8.5 to 105 days — hair, averaging a 7 to 10 day lag before use grows out and a window running from 90 days to six months depending on the source

The blood figures are the weakest part of this page. None of the ten sources gives a per-drug blood window - they report a general one to two days across drugs - so the blood row does not change with frequency, and it would be inventing precision to make it. Urine is where these sources are strong. And the tails run past every upper bound here: both StatPearls chapters report cannabis at three months after heavy chronic use.

The ten sources

Every window on this page is the average of what these ten sources report. Nine are US federal or National Institutes of Health publications and one is a university public-health department - no commercial treatment-centre pages, which is what an earlier version of this page leaned on. The per-drug notes under the results table say how many of the ten actually cover that substance, because several cover only some of them.

How the average is taken

low  = mean of every source's lower bound
high = mean of every source's upper bound

Sources that report only "up to X" contribute to the upper average and are left out of the lower one, rather than being counted as a zero. Cannabis is the only substance the sources band by how often you used, so its five bands are each averaged separately. For everything else the sourced average is the regular-use band and the others are scaled from it, which the working section sets out in full.

SAMHSA. Medications for Opioid Use Disorder, TIP 63 - Urine Drug Testing Window of Detection. Federal. The most complete per-drug table of the ten, and the only source here that breaks out oxycodone and hydrocodone separately.
SAMHSA. Substance Abuse: Clinical Issues in Intensive Outpatient Treatment, TIP 47, Appendix B. Federal. Bands cannabis into casual, daily and chronic, and is one of only two sources here giving ethanol, blood and hair figures.
Drug Testing. StatPearls, NCBI Bookshelf, National Library of Medicine. Source of the heavy-chronic figures - amphetamines 10 days, cocaine metabolites three weeks, heroin metabolites 11 days, cannabinoids three months.
Clinical Drug Testing. StatPearls, NCBI Bookshelf, National Library of Medicine. Corroborates the same heavy-use set independently, and gives the plasma window of one to two days.
Objective Testing - Urine and Other Drug Tests. PubMed Central, National Institutes of Health. Table 1 covers urine, hair, oral fluid and sweat together, and is the only source here that quantifies an EtG alcohol window.
Interpretation of Workplace Tests for Cannabinoids. PubMed Central, National Institutes of Health. The federal cannabinoid cutoffs of 50 ng/mL to screen and 15 ng/mL to confirm, and the finding that the metabolite persists for weeks or months in chronic users.
Urine drug screens: considerations for the psychiatric pharmacist. PubMed Central, NIH. The 6-MAM window of under 12 hours, and the general finding that most drugs are detectable for one to three days.
Verstraete AG. Detection Times of Drugs of Abuse in Blood, Urine, and Oral Fluid. Therapeutic Drug Monitoring 26(2), 2004. Peer-reviewed. Single dose in urine 1.5 to 4 days, chronic users about a week and longer for cocaine and cannabis, blood one to two days.
NYU School of Global Public Health. Drug and Alcohol Testing. Academic. The hair figures - detection starting 7 to 10 days after use, covering roughly the past 90 days.
SAMHSA. Frequently Asked Questions About Federal Workplace Drug Testing. Federal. Which specimen types are actually authorised for regulated testing, and the cutoffs those programmes use. Hair is still not an approved federal matrix.

One more, which is not part of the average because it does not report detection windows. It is the reason body weight is asked for and then not used: Schuster RM et al. Urinary 11-nor-9-carboxy-tetrahydrocannabinol elimination in adolescent and young adult cannabis users during one month of sustained and biochemically-verified abstinence. J Psychopharmacol 34(2), 2019. It followed 62 users through a verified month of abstinence and found sex, BMI, race, years of use and past-month frequency were all not significantly associated with clearance rate. It also puts the median detection window at 10 days with a 95% prediction interval of 4 to 80 days, which is a blunter statement of the spread than anything else on this page - and 40% of participants still tested positive after 25 dry days.

Hair is now sourced rather than assumed. Two sources give the same 7 to 10 day lag before use grows far enough out of the scalp to be sampled, which is where the 8.5-day lower bound comes from. Three put the window near 90 days and SAMHSA TIP 47 puts it at four to six months, which is why the average lands at 105. Worth knowing that hair is still not an approved matrix for federally regulated testing at all.