How to test cortisol levels: what each test actually checks

How to test cortisol levels: the main options

If you want to know how to test cortisol levels, the short answer is there is no single test. Doctors choose from a morning blood draw, a 24-hour urine collection, a late-night saliva sample, or a drug-challenge test like the dexamethasone suppression test or the ACTH stimulation test, depending on what they suspect.

Each method measures something slightly different, and each has its own quirks. A blood test captures cortisol at one moment. Urine and saliva tests can capture a total or a specific trough. The suppression and stimulation tests check whether your body's cortisol system responds the way it should when pushed, rather than just reporting a number. None is inherently "better." The right one depends on whether a doctor is screening for too much cortisol, as in Cushing syndrome, or too little, as in Addison's disease.

Why a single blood draw can mislead you

Cortisol does not sit still. According to Cleveland Clinic, blood levels typically run 10 to 20 mcg/dL between 6 and 8 a.m. and fall to roughly 3 to 10 mcg/dL by mid-afternoon. That is not noise. It is the normal diurnal rhythm, the same daily pattern behind the morning cortisol spike that helps get people out of bed.

A single draw taken at the wrong time, or on a stressful day, can look abnormal in a perfectly healthy person. MedlinePlus notes that stress, pregnancy, exercise, temperature extremes, and medications including birth control pills can all shift a result. Even the needle stick itself is a minor stressor that can nudge cortisol upward. That is part of why Cleveland Clinic says you may need to be tested twice in one day, or on separate days, before anyone draws a conclusion.

The 24-hour urine and late-night saliva tests

Two of the three tests the Endocrine Society's guideline recommends as a first screen for Cushing syndrome sidestep the single-moment problem entirely. A 24-hour urine free cortisol test has you collect every bit of urine over a full day, which adds up total cortisol output rather than catching one instant. MedlinePlus describes the collection: discard the first morning void, keep the container cold, and catch everything for the next 24 hours.

Late-night salivary cortisol works differently. Cortisol is supposed to be near its lowest around midnight, so a swab taken then should read low. In Cushing syndrome, it typically does not. NICHD notes the sample can be collected with a special device at home and mailed to a lab, which makes it one of the more practical outpatient options. Both tests still depend on normal conditions: shift work, irregular sleep, or a late workout can blur the result.

Drug challenge tests: dexamethasone suppression and ACTH stimulation

The dexamethasone suppression test flips the question around. You take a dose of the steroid dexamethasone, which should tell a healthy pituitary gland to stop signaling the adrenal glands to make cortisol. NICHD describes an overnight version, with a dose at 11 p.m. and a blood draw the next morning, plus a longer three-day protocol. If cortisol does not drop as expected, that points toward Cushing syndrome.

The ACTH stimulation test works in the opposite direction and is used for suspected adrenal insufficiency, not excess. NIDDK explains that a clinician injects synthetic ACTH, the hormone that normally tells the adrenal glands to produce cortisol, then draws blood before the injection and again 30 to 60 minutes later. A healthy adrenal gland responds with a jump in cortisol. Little or no rise suggests Addison's disease, though NIDDK notes the test can look falsely normal very early in some pituitary-related cases, before the adrenal glands have had time to shrink.

Hair cortisol: measuring the long view

Hair cortisol testing is pitched as a way to see months of cortisol exposure instead of one day, since hair grows at a fairly steady rate and effectively banks a record as it goes. The idea has real research behind it, including small studies finding decent short-interval reliability. The evidence is still thinner than for the tests above, and this piece should say so plainly rather than oversell it.

Studies have struggled with inconsistent lab methods between sites, and at least one found hair cortisol did not line up well with same-period urine cortisol in the same people. It has not replaced the standard Cushing's or Addison's workups in clinical guidelines. Right now it is better understood as a research tool than a diagnostic one you should expect your doctor to order.

What counts as a "normal" cortisol level

Even within a single test type, "normal" shifts with the clock and the lab. Cleveland Clinic lists roughly 10 to 20 mcg/dL for a 6 to 8 a.m. blood draw and 3 to 10 mcg/dL around 4 p.m. Healthline cites a slightly wider 8 a.m. range of 5 to 25 mcg/dL. Neither is wrong. They reflect different reference populations and assay methods, which is exactly why MedlinePlus tells patients to read the range printed on their own lab report rather than a number pulled from the internet.

This is also why doctors rarely judge a cortisol result in isolation. A urine or saliva result gets compared against a lab's own cutoff for that collection type and time. A suppression or stimulation test is judged by how much cortisol changes, not by a single absolute value. The number alone, without the context of which test, what time, and what the body was asked to do, tells you very little on its own.

At-home cortisol test kits: a screen, not a diagnosis

Direct-to-consumer kits have multiplied alongside interest in stress hormones generally, mostly using saliva swabs or finger-prick blood spots you mail to a lab. Healthline's review of the category found that not all of them are reliable, and that testing at a doctor's office or hospital is generally more accurate than a self-collected sample, in part because timing and handling errors are easy to make at home.

That does not make home kits worthless. A kit from a company that names a CLIA-certified lab and gives you a clear collection window can show you a rough diurnal pattern worth discussing with a doctor. What it cannot do is diagnose Cushing syndrome or Addison's disease on its own. An unusual result is a reason to ask a clinician for the clinic-supervised version of the same test, not a verdict in itself.

When a doctor actually orders these tests

Guidelines are specific about who needs cortisol testing in the first place. A 2 Minute Medicine summary of the Endocrine Society's recommendations notes that screening is appropriate once exogenous steroid use is ruled out and a patient has multiple progressive, clinically compatible features, not an isolated complaint about a puffy face or everyday stress. Testing for vague, non-specific stress is explicitly discouraged, partly because untargeted testing produces false positives in people with obesity, PCOS, depression, or poorly controlled diabetes.

In practice, that means these tests show up when someone has a cluster of symptoms: unexplained weight gain with a rounded face and new bruising for suspected Cushing's, or persistent fatigue, weight loss, and low blood pressure for suspected Addison's. Feeling stressed is not, on its own, an indication for any of the tests above.

Our take

Testing cortisol is genuinely useful when a doctor is working through a real suspicion of Cushing syndrome or Addison's disease, and the tests above are good at answering that specific question. It is a much weaker tool for curiosity about everyday stress, since a single number taken at the wrong time can easily look abnormal in a healthy person. If you are going to test, match the test to the question and let a clinician read the result in context.

This is general information, not medical advice. Talk to a clinician before changing anything, especially if you are pregnant, nursing, or managing a condition.