How to Check Cortisol Levels: Blood, Saliva and Urine Tests Compared

The video lasts nine seconds. A woman presses her fingertips to her cheeks, mouths the words “cortisol face” and swipes to a before-and-after. Underneath, thousands of comments ask the same thing: how do I get mine tested? Clips like that explain why searches for how to check cortisol levels climbed through 2025 and, as of early 2026, remain near their peak, carried along by a parallel boom in mail-in saliva kits that promise a stress “readout” within a week.

Here is the quiet problem. Cortisol is one of the most time-sensitive measurements in the laboratory. The same person can post a value at 8 a.m. that is several times higher than the one at midnight, and both are normal. A test that ignores the clock, or the inhaler in your bathroom cabinet, can look alarming while telling you nothing.

So this guide does what the viral clips do not. It walks through the three sample types clinicians actually use, what each is good at, what trips each one up, and what a result can and cannot say about your stress.

What cortisol does, and why one number rarely tells the story

Cortisol is a steroid hormone made by the adrenal glands, two walnut-sized organs perched on top of the kidneys. Its reputation as “the stress hormone” undersells it. Cortisol raises blood sugar between meals, keeps blood pressure from sagging when you stand, damps inflammation and helps you wake up. Without it, the body cannot cope with a fever, a fast or a surgery.

Production is run by a chain of command called the HPA axis: the hypothalamus in the brain signals the pituitary gland, the pituitary releases ACTH (adrenocorticotropic hormone, the messenger that tells the adrenals to work), and the adrenals respond with cortisol. Rising cortisol then feeds back to quiet the first two links. The loop is why a result is often paired with an ACTH measurement; the two together say where a problem sits.

Two features of this system matter for testing. First, cortisol follows a diurnal rhythm, meaning a built-in daily cycle: it surges in the first hour after waking, drifts down through the afternoon and reaches its floor around midnight. Second, roughly nine-tenths of the cortisol in blood rides attached to carrier proteins, mainly cortisol-binding globulin; only the unattached “free” fraction is biologically active. Blood tests count both fractions, while saliva and urine tests count mostly the free portion.

Put those facts together and the limits of a single number become obvious. A value with no time stamp is unreadable. A total blood level in someone taking an estrogen-containing pill may look high purely because the carrier protein has risen. And a stressful morning, including the needle itself, can push a reading up without anything being wrong. Clinicians therefore ask three questions of every result: when was it taken, which fraction did it measure, and what else was going on in the body that day.

Why is it important to check cortisol levels, and when is it not?

In clinical practice, cortisol is measured to answer a narrow set of questions, not to grade how stressed someone feels. The first is whether the body is making far too much, a condition called Cushing’s syndrome. It is rare: the National Institute of Diabetes and Digestive and Kidney Diseases estimates roughly 40 to 70 new cases per million people each year, and most of those are caused by steroid medicines rather than a tumor. The second question is whether the body is making too little, called adrenal insufficiency; Addison’s disease is the form where the adrenal glands themselves are damaged. The third is whether long-term steroid treatment has switched off the body’s own production, which matters when a medicine is being tapered.

Symptoms of high cortisol tend to accumulate over months rather than appear overnight. Weight settles around the trunk, face and the back of the neck while arms and legs stay thin. Wide purple stretch marks may appear on the abdomen, skin bruises with little provocation, and blood pressure or blood sugar creep up. Muscles in the thighs weaken, so rising from a chair becomes effortful. Mood changes and irregular periods are common.

Low cortisol looks almost like the mirror image: persistent exhaustion, weight loss, nausea, salt craving, dizziness on standing and, in Addison’s disease, a tan-like darkening of skin creases and gums.

What is not on either list is ordinary, explicable stress. Feeling frazzled after a hard quarter at work does not make a cortisol test informative, and no major guideline recommends routine cortisol screening in healthy adults. The reason is practical: in a person with no physical signs, an abnormal-looking number is more likely to reflect timing, medicines or lab variation than disease, and chasing it can lead to unnecessary scans. Testing earns its place when symptoms point toward too much or too little hormone, when a steroid medicine is being adjusted, or when an incidental adrenal finding on a scan needs explaining.

What changed recently

The science of cortisol testing has moved slowly; the conversation around it has not. Three shifts explain the current surge.

The first is cultural. Across 2024 and 2025, short-form videos attributing facial puffiness, abdominal weight and poor sleep to “high cortisol” drew hundreds of millions of views, and search interest in cortisol testing rose with them. None of those clips changed the medical picture, but they changed who shows up asking for a test.

The second is commercial. Direct-to-consumer kits that once offered only a morning saliva sample now sell four-point daily profiles, dried blood spot cards and, increasingly, hair analysis. The technology in the better kits is the same immunoassay chemistry hospital labs use. What differs is the lack of a clinician choosing the timing and interpreting the result against your history.

The third shift is clinical and dates back further than the trend suggests. In 2008, the Endocrine Society’s practice guideline formalized three first-line screening tests for suspected Cushing’s syndrome: late-night salivary cortisol, 24-hour urine free cortisol and a low-dose dexamethasone suppression test. That trio remains the standard described on current patient pages from the Mayo Clinic and the NIDDK, with late-night saliva having moved from a research tool to a routine one over the intervening years because it is painless and can be collected at home.

Two quieter developments deserve mention. Hair cortisol, which averages exposure over roughly a month per centimeter of growth, has accumulated hundreds of PubMed-indexed studies in the past decade and is now a recognized research measure of chronic exposure, though not a diagnostic test. And endocrinology services report more cases of adrenal suppression from steroid creams, inhalers and joint injections, which has pushed labs to ask patients about every steroid-containing product, not just tablets, before a sample is drawn.

How to check cortisol levels with a blood test

The blood test is the one most people picture. A sample is drawn from a vein in the arm, almost always in the morning, when cortisol is near its peak; some clinicians add a second draw in the late afternoon to see whether the expected decline has happened. The lab reports serum cortisol, which means total cortisol, bound and free together.

Where it shines is in suspected low cortisol. A clearly low morning value in someone with matching symptoms is a strong signal, and a clearly normal-to-high one makes adrenal insufficiency unlikely. The blood sample also serves as the baseline for the dynamic tests described later, and it can be drawn alongside ACTH, electrolytes and glucose, which often tell part of the same story.

Where it struggles is as a stand-alone screen for too much cortisol. Secretion comes in pulses, so two draws ten minutes apart can differ meaningfully. A single morning value sits in the normal range for many people with early Cushing’s syndrome, because the hallmark of the disease is not a towering peak but a failure to fall at night, and a morning sample never sees the night. The venipuncture itself can nudge the number upward in anyone anxious about needles.

Preparation is simple but not trivial. You will usually be asked to arrive within an hour or so of your normal waking time, to avoid strenuous exercise beforehand, and to list every medicine and supplement, including inhalers, nasal sprays, skin creams and anything containing biotin, which can interfere with the chemistry of some assays. Results typically return within a few days.

One more nuance: because the test counts bound cortisol, anything that raises cortisol-binding globulin, such as pregnancy or estrogen-containing contraception, raises the total without changing the active fraction. A clinician who knows this will either interpret accordingly or reach for a saliva or urine test instead.

The saliva cortisol test: why late night beats morning

Saliva has a specific advantage: it contains almost no binding protein, so what you spit into the tube is free cortisol, the fraction that actually acts on tissues. Collection is painless and can be done at home. You either let saliva pool and drool it through a straw into a vial or hold an absorbent swab in the mouth for a minute or two, then refrigerate or mail the sample as instructed.

Timing is everything. The clinically useful saliva test is the late-night one, usually collected between about 11 p.m. and midnight, because that is when cortisol should be at its lowest. People with Cushing’s syndrome lose that midnight trough early in the disease, often before anything else is abnormal. Diagnostic-accuracy studies in referral populations have reported sensitivity and specificity above 90 percent for a late-night salivary cortisol, which is why the Mayo Clinic and the NIDDK list it among the first-line screens. Clinicians typically request two or more nights, since a single elevated sample can reflect a late dinner or a bad day.

The test is far less helpful for diagnosing low cortisol, where blood remains the standard, and it is nearly uninterpretable in people whose rhythm is not aligned to the clock. Night-shift workers, new parents feeding every two hours and recent long-haul travelers have real “midnight” values that mean something different.

Small details corrupt the sample. Eating, drinking anything other than water, brushing teeth or smoking in the half hour before collection can contaminate it, and bleeding gums add blood cortisol to the saliva, inflating the result. Steroid creams on the hands can contaminate the swab. Most labs send printed instructions; following them to the letter matters more here than with any other sample type.

Many wellness kits also sell four-sample daytime “curves.” These show the shape of your rhythm, which can be interesting, but the daytime points are not validated for diagnosing disease the way the late-night point is.

The 24-hour urine test: measuring the whole day at once

If the blood test is a snapshot and the saliva test a well-timed photograph of the night, the 24-hour urine test is a long exposure. You discard your first morning urine, then collect every drop for the next 24 hours into a large container kept cool, finishing with the first void the following morning. The lab measures urinary free cortisol, the unbound hormone that spilled into the urine across the entire day, and usually checks creatinine, a muscle waste product excreted at a steady rate, to confirm the collection is complete.

That integration is the test’s strength. Momentary pulses and needle anxiety wash out when a whole day is averaged, and a value several times the upper limit of the reference range points firmly toward excess production. The test has been used for decades and remains a first-line screen for suspected Cushing’s syndrome.

Its weaknesses are mostly human. Missing a single bathroom visit, perhaps the one at work, undercounts the total and can mask a problem. Drinking very large volumes of fluid, more than about five liters a day, increases cortisol excretion and can falsely elevate the result. Reduced kidney function pushes values down. Mild or cyclical Cushing’s syndrome can produce a normal urine result while a late-night saliva sample is abnormal, so most clinicians order two collections and rarely rely on urine alone.

The urine test is not used to diagnose low cortisol, because the lower end of the normal range overlaps too much with disease. Nor can it be collected casually; the container, preservative and timing instructions come from the lab, and the sample needs to be returned promptly.

For a person who dreads needles and whose work schedule makes a midnight saliva collection unreliable, the urine test is often the most practical of the three. For someone with an irregular bladder routine or a demanding commute, it is the easiest to get wrong.

Dynamic tests: when a single sample is not enough

Static measurements tell you how much cortisol is present. Dynamic tests ask how the system responds when pushed, and they are where diagnosis usually gets settled. Both are ordered and supervised by a clinician; the details of timing and quantity follow laboratory protocols and are not something to arrange yourself.

The dexamethasone suppression test uses dexamethasone, a synthetic steroid that the body’s feedback loop recognizes as cortisol. In a healthy person, a small amount taken under the protocol tells the pituitary to stop sending ACTH, and the next morning’s blood cortisol falls to a low level. In Cushing’s syndrome, production carries on regardless and the morning value stays up. The test is a first-line screen for excess cortisol. It can be confounded by medicines that speed the breakdown of dexamethasone, including some anti-seizure drugs, and by the same binding-protein effects that complicate any blood measurement.

The ACTH stimulation test runs in the opposite direction and is used when too little cortisol is suspected. A baseline blood sample is taken, a synthetic form of ACTH is injected, and blood is drawn again after a set interval. Healthy adrenal glands respond with a brisk rise; damaged or long-suppressed glands do not. The NHS describes this as the standard way to confirm Addison’s disease after a low morning cortisol has raised suspicion.

Once an abnormality is confirmed, an ACTH level helps locate the cause. High cortisol with low ACTH suggests the adrenal gland itself; high cortisol with high ACTH points toward the pituitary or, rarely, a tumor elsewhere making ACTH. Only then do imaging scans enter the picture.

The sequence matters because scans alone mislead. Small, harmless adrenal and pituitary nodules are common incidental findings, and imaging a person before biochemistry has shown a real excess risks blaming an innocent bystander.

Blood vs saliva vs urine: how the three cortisol tests compare

Each sample type answers a slightly different question, and the best choice depends on whether a clinician suspects too much hormone, too little, or a disrupted rhythm. The table below summarizes the trade-offs that come up most often in clinic.

Feature	Blood (serum)	Saliva (late night)	Urine (24-hour)
What it measures	Total cortisol, bound plus free	Free cortisol at one moment	Free cortisol averaged over a day
Usual timing	Morning, sometimes late afternoon too	Around 11 p.m. to midnight, two or more nights	Full 24-hour collection, often repeated
Best for	Suspected low cortisol; baseline for dynamic tests	Screening for Cushing’s syndrome; loss of the night-time trough	Screening for Cushing’s syndrome; integrated excess
Weak for	Detecting mild excess from one sample	Diagnosing low cortisol; shift workers	Diagnosing low cortisol; mild or cyclical excess
Main confounders	Estrogen, pregnancy, biotin, needle stress	Food, gum bleeding, steroid creams, irregular sleep	Missed voids, very high fluid intake, kidney function
Where collected	Clinic or lab	Home	Home, container from lab
Discomfort	Needle	None	None, but inconvenient

Reading across the rows reveals a pattern that surprises many patients: the test that feels most medical, the blood draw, is the least reliable way to catch a mild excess, while the test that can be done in pajamas is among the most sensitive for it. Conversely, nobody should accept a diagnosis of low cortisol on a saliva kit alone.

In real practice the three are rarely rivals. A typical work-up for suspected excess pairs two late-night saliva samples with a suppression test or a urine collection, and only when at least two independent tests agree does the clinician move to ACTH levels and imaging. For suspected deficiency, a morning blood value opens the door and a stimulation test confirms. The strongest results come from asking the same question two different ways.

Normal cortisol levels by time of day: reading the ranges

Patients often open a results portal, see a number flagged in orange and start searching for a single “normal cortisol level.” There isn’t one. There are several, and they belong to specific hours.

The daily curve runs like this. Cortisol begins rising in the last hours of sleep and jumps sharply in the 30 to 45 minutes after waking, a surge called the cortisol awakening response that typically lifts the level by about half. From that peak it declines in steps through the day, with small bumps after meals and exertion, and by late evening it has fallen to a small fraction of the morning value. The midnight floor is the point disease tends to disturb first.

Because of that shape, laboratories print separate reference ranges for morning and afternoon blood samples, a separate and much lower range for late-night saliva, and a per-day range for urine. Units add a second layer of confusion: North American labs usually report blood cortisol in micrograms per deciliter and many others in nanomoles per liter, and the two scales differ by a factor of about 27.6. A number that looks enormous in one unit can be ordinary in the other.

Reference ranges also differ between laboratories because assays differ. Immunoassays, which detect cortisol with antibodies, can cross-react with related steroids and tend to read slightly higher than mass spectrometry, a method that identifies the molecule by weight. The range printed beside your result is the only one that applies to it; a range quoted on a website or from a different lab is not comparable.

Age and sex change the ranges little in adults, though pregnancy roughly doubles total blood cortisol by the third trimester, and newborns have not yet developed a day-night rhythm at all. For practical purposes, the question to ask a clinician is not “is my number normal?” but “is my number normal for the time, sample and method used?”

Can you check cortisol levels at home? Kits, hair tests and wearables

A cortisol test at home is now easy to buy, and the honest answer about its value is: it depends which part of the medical process the kit replaces. Two parts matter, the sample and the interpretation. Home kits can handle the first reasonably well and the second poorly.

Saliva kits are the most defensible. The late-night salivary cortisol that endocrinologists rely on is already a home-collected sample, and reputable mail-in labs run the same immunoassays. What the consumer version lacks is a clinician deciding whether the test is appropriate for you, specifying how many nights to collect, screening for confounders such as steroid creams or shift work, and placing the result beside your symptoms. A single borderline value with no context is more often a source of anxiety than information. Daytime “stress curve” kits add four or five extra samples that are interesting to look at but not validated to diagnose anything.

Dried blood spot cards, where you prick a finger and let drops dry on filter paper, measure total cortisol much as a vein sample does, with the same timing and binding-protein caveats and a little more variability from collection technique.

Hair cortisol is different in kind. A centimeter of scalp hair holds roughly a month of cortisol exposure, so a three-centimeter sample gives a seasonal average. Research groups use it to study chronic stress at the population level, and the correlations are real. For an individual, however, hair-washing habits, dyes, hair type and the assay itself add enough noise that no clinical guideline accepts it for diagnosis. It is a research measure, not a test to act on.

Then there are devices that are not cortisol tests at all. A continuous glucose monitor and the apps built around it measure sugar in the fluid under the skin, not cortisol; a glucose curve can rise after a stressful event because cortisol liberates stored sugar, but that is an indirect echo, not a measurement. Smartwatches estimate “stress” from heart rate variability, which reflects the nervous system’s balance and has no fixed relationship to cortisol. Useful tools, wrong question.

What can skew a cortisol result before you even reach the lab

Most surprising cortisol results have an unglamorous explanation, and a careful clinician will hunt for it before ordering anything else. The list is longer than many people expect.

Steroid medicines head it. Prednisone, hydrocortisone and related drugs are recognized by many assays as cortisol and push readings up; over weeks they also suppress the body’s own production, so a reading taken after a course has ended can be low. The products patients forget to mention cause the most confusion: inhalers for asthma, nasal sprays for allergies, eczema creams, hemorrhoid preparations and steroid joint injections all contain enough to matter. Dexamethasone is the exception that cuts the other way, since most assays do not detect it but it still switches off natural cortisol, which is exactly why it is used in the suppression test.

Hormones come next. Estrogen-containing contraception and pregnancy raise cortisol-binding globulin, lifting total blood cortisol without changing the free fraction; saliva and urine tests sidestep this. Severe illness, recent surgery, uncontrolled pain, poorly controlled diabetes, heavy alcohol use, significant depression and severe obesity can all raise cortisol modestly through the HPA axis itself, a pattern sometimes called pseudo-Cushing’s because it mimics early disease on paper without the physical signs.

Timing errors are the commonest of all. A “morning” sample drawn at 11 a.m. from someone who woke at 5 is really a late-morning sample. A late-night saliva sample from a nurse coming off a night shift is a morning sample in disguise. Crossing several time zones in the preceding week shifts the entire curve.

Finally, the chemistry. High-intake biotin supplements, marketed for hair and nails, can interfere with some immunoassays and distort results in either direction, which is why many labs ask you to pause them before testing. Bleeding gums contaminate saliva, incomplete collections sabotage urine, and samples left warm in a mailbox degrade. None of these is exotic; together they account for a large share of the abnormal results that turn out to mean nothing.

What the evidence actually says about cortisol testing

Not all claims in this field rest on the same foundation, so it helps to grade them.

The daily rhythm of cortisol, and the physiology of the HPA axis behind it, is as settled as anything in endocrinology. It rests on decades of repeated sampling studies in humans and on the predictable consequences of removing or over-stimulating each gland in the chain. Treat this as established fact.

The diagnostic performance of the three screening tests for Cushing’s syndrome is well supported. The relevant evidence is not randomized trials, which are the wrong tool for a question about test accuracy, but diagnostic-accuracy cohorts comparing each test against confirmed diagnoses in referral populations. Those studies consistently show high sensitivity and specificity for late-night salivary cortisol and the dexamethasone suppression test, and somewhat lower sensitivity for 24-hour urine in mild disease. The main caveat is that performance in referral clinics is better than in the general population, where disease is rarer and false positives proportionally more common. Grade this as strong, with the usual limits of observational data.

The use of a morning blood cortisol followed by an ACTH stimulation test for suspected adrenal insufficiency is similarly well supported by observational studies and long clinical use, and is described as standard by the NHS and NIDDK.

Hair cortisol as a marker of chronic stress sits at moderate strength for group-level research and weak strength for individual interpretation; correlations with reported stress are real but modest, and pre-analytical variables are large.

Evidence that checking cortisol in healthy, stressed people improves any outcome is essentially absent. No trial has shown that measuring cortisol changes health in people without signs of disease. The concept of “adrenal fatigue” was examined in a 2016 systematic review of 58 studies indexed in PubMed, which found no consistent cortisol pattern and concluded the condition lacks evidence. Claims for supplements that “lower cortisol” rest on small, short studies with mixed results and no demonstrated clinical benefit.

Common myths about cortisol face, cortisol belly and adrenal fatigue

Viral claims about cortisol tend to borrow a real fact and stretch it past breaking point. A few deserve direct correction.

“Facial puffiness means high cortisol.” Genuinely excessive cortisol, as in Cushing’s syndrome, does round the face, but it does so alongside purple stretch marks, central weight gain, thinning skin, muscle weakness and rising blood pressure, over months. Ordinary morning puffiness from salt, alcohol, poor sleep or allergies is fluid, not hormone, and a cortisol test in that situation is far more likely to be normal than revealing.

“Belly fat is cortisol fat.” Cortisol does favor fat storage around the trunk, which is why it features in Cushing’s syndrome. In the general population, abdominal fat tracks with total calories, inactivity, genetics, age and sleep far more than with measurable cortisol differences, and lowering cortisol is not a validated route to losing it.

“You probably have adrenal fatigue.” The idea that chronic stress “exhausts” the adrenal glands into underproduction is not recognized by endocrine societies and was not supported by a systematic review of the literature. Real adrenal insufficiency exists, is diagnosable with the tests above, and is medically serious, which is exactly why a vague label that bypasses testing is unhelpful.

“One morning blood test rules it in or out.” For low cortisol, a morning sample is a reasonable start. For high cortisol, it is close to useless on its own, because early disease shows up at midnight, not at breakfast.

“You can detox cortisol.” Cortisol is not a toxin and does not accumulate to be flushed. It is cleared continuously by the liver and kidneys, and healthy sleep, activity and stress management influence its rhythm through the brain, not through any cleanse.

“A home kit can diagnose Cushing’s.” A home saliva sample can be one piece of a diagnosis made by a clinician. No single test, at home or in hospital, diagnoses Cushing’s syndrome by itself.

When to see a doctor about cortisol

Most people who wonder about their cortisol do not need a test; some do, and a few need urgent care. Knowing which group you are in matters more than knowing which kit to buy.

Book a routine appointment if, over weeks to months, you have noticed several of the following together: weight gain concentrated on the trunk, face or upper back while the limbs stay slim; wide purple or red stretch marks; skin that bruises or tears easily; new or worsening high blood pressure or blood sugar; thigh weakness that makes stairs or standing from a chair hard; or irregular periods with new facial hair. The combination, not any one item, is what prompts testing.

Also book if you have the opposite cluster: persistent exhaustion that sleep does not fix, unintended weight loss, nausea or reduced appetite, salt craving, dizziness or faintness on standing, or a tan-like darkening of skin creases, scars or gums.

Seek emergency care without delay for severe vomiting or diarrhea with weakness, confusion, severe abdominal or back pain, or a collapse, particularly in anyone who takes steroid medicines, has recently stopped or tapered them, or has a known adrenal or pituitary condition. These can signal adrenal crisis, a life-threatening shortage of cortisol that needs immediate treatment.

Children who gain weight while their growth slows should be assessed, since that pairing is a classic sign of excess cortisol in childhood.

Two rules apply to everyone. Never stop, skip or reduce a prescribed steroid medicine because of a cortisol result or a video; sudden withdrawal after prolonged use can itself trigger adrenal crisis, and any change belongs to the prescriber. And bring every result, including home kits, to the clinician who will interpret it, along with a complete list of medicines, inhalers, creams, supplements and your usual sleep schedule. Decisions about further testing, imaging or treatment rest with that clinician, who can see the whole picture that a single number cannot.

Frequently asked questions

How can I check cortisol levels at home?

The only home-collected cortisol measurement clinicians routinely use is a late-night saliva sample, taken around 11 p.m. to midnight on two or more nights. Mail-in kits can run the same assay a hospital lab uses, but they do not decide whether testing is appropriate for you or interpret the result against your medicines and symptoms. Hair tests and dried blood spots exist but are research tools or less reliable versions of lab tests. Take any home result to a clinician rather than acting on it.

Why is it important to check cortisol levels?

Testing matters when symptoms suggest the body is making far too much cortisol, as in Cushing’s syndrome, or too little, as in adrenal insufficiency, and when a long-term steroid medicine is being tapered. Both conditions are treatable, and the second can be dangerous if missed. Feeling stressed on its own is not a reason to test; no major guideline recommends routine cortisol screening in healthy adults, because out-of-context numbers mislead more often than they help.

Is a glucose-monitoring app a cortisol test?

No. Continuous glucose monitors and the metabolic-health apps built around them measure sugar in the fluid just under the skin, not cortisol. Because cortisol raises blood sugar, a glucose line may bump upward after a stressful event, but that is an indirect effect shared with meals, exercise and illness. A glucose trace cannot tell you whether your cortisol is normal, and it plays no role in diagnosing Cushing’s syndrome or adrenal insufficiency.

How does a CGM-based wellness app work, and does it show stress hormones?

A small sensor worn on the arm or abdomen samples glucose from tissue fluid every few minutes and sends readings to a phone app, which charts how food, sleep and activity move the curve. These systems were designed for diabetes management and are now marketed for general metabolic insight. They do not measure cortisol, adrenaline or any hormone directly, so any “stress” label an app applies is an inference from glucose and heart-rate patterns, not a measurement.

How accurate is a saliva cortisol test?

For its intended purpose, screening for excess cortisol, a properly collected late-night salivary cortisol performs well: diagnostic-accuracy studies in specialist clinics have reported sensitivity and specificity above 90 percent. Accuracy drops sharply when timing is wrong, when the sample is contaminated by food, blood from gums or steroid creams, or when the person works nights. It is not an accurate way to diagnose low cortisol, where a morning blood test and a stimulation test remain the standard.

Do I need to fast before a cortisol blood test?

Usually not, though your lab may ask you to avoid food if other tests, such as glucose or lipids, are drawn at the same time. What matters more is timing and preparation: arriving at the requested morning slot, avoiding hard exercise beforehand, resting a few minutes before the draw, and telling staff about every medicine, inhaler, cream and supplement you use, including biotin, which can interfere with some assays. Follow the specific instructions your clinician or lab provides.

What time of day are cortisol levels highest and lowest?

Cortisol is highest in the first hour after waking, rising by roughly half in the 30 to 45 minutes after you open your eyes, and lowest around midnight. Between those points it declines in steps, with small rises after meals and exertion. Laboratories print separate reference ranges for morning, afternoon and late-night samples because of this curve, so a value can only be judged against the range for the hour it was collected.

Can stress alone cause an abnormal cortisol test?

It can push a result toward or just past the upper limit, especially a single morning blood sample taken from someone anxious about the needle. Severe depression, heavy alcohol use, acute illness and poorly controlled diabetes can raise cortisol more persistently, a pattern sometimes called pseudo-Cushing’s. Clinicians distinguish this from true excess by repeating tests, using more than one sample type and looking for physical signs such as stretch marks, bruising and muscle weakness.

Does the contraceptive pill affect cortisol results?

Estrogen-containing contraceptives raise cortisol-binding globulin, the carrier protein in blood, so a serum cortisol can read high even though the active free fraction is unchanged. This can produce a false positive on blood-based tests, including the dexamethasone suppression test. Saliva and 24-hour urine tests measure free cortisol and are largely unaffected. Tell your clinician about any hormonal contraception before testing; they may choose a different sample type or interpret accordingly. Do not stop contraception on your own.

How long does it take to get cortisol test results, and what happens if they are abnormal?

Blood results commonly return within a few days; saliva and urine samples sent to a reference lab may take a week or more. One abnormal screening result rarely leads straight to a diagnosis. The usual next step is a second, different test, since clinicians want at least two independent measurements to agree before ordering ACTH levels or imaging. Many initially abnormal results normalize on repeat once timing, medicines and collection problems are addressed.

References

• MedlinePlus: Cortisol Test
• NHS: Addison's disease, diagnosis
• Cleveland Clinic: Cortisol
• NIH NIDDK: Cushing's Syndrome

This article is for general information only and is not a substitute for professional medical advice. Please consult a qualified doctor about your individual situation.