Cortisol testing comes in several forms, and the right one depends on what your doctor is looking for. A morning blood draw measures total cortisol circulating in your bloodstream, a saliva sample captures the free (biologically active) fraction, and a 24-hour urine collection reflects your cumulative cortisol output over a full day. Each method answers a slightly different clinical question, and each has quirks that can throw off results if you are not prepared for them.
What a Blood Test Actually Measures
When most people hear “cortisol test,” they picture a standard blood draw. A nurse takes a sample, usually between 7 and 9 a.m. when cortisol peaks, and the lab returns a number in micrograms per deciliter (or nanomoles per liter, depending on the country). That number represents total serum cortisol, which is the sum of two fractions: the cortisol bound to a carrier protein called corticosteroid-binding globulin (CBG) and the small unbound portion that is actually active in your tissues.
This distinction matters more than it might sound. Roughly 80 to 90 percent of the cortisol in your blood is bound to CBG and is essentially inactive. So the total number on your lab report is heavily influenced by how much CBG you have, not just how much cortisol your adrenal glands are producing. Anything that raises or lowers CBG will shift the total cortisol reading even if your body’s actual cortisol activity has not changed. Oral estrogen, whether from contraceptive pills or hormone therapy, is the most common culprit: it raises CBG substantially, which in turn inflates the total cortisol reading.1PubMed. Prospective evaluation of the effect of stopping estrogen-containing contraceptives on serum CBG and cortisol levels Current guidelines recommend stopping estrogen-containing contraceptives at least six weeks before cortisol blood testing to avoid misleading results.
Pregnancy does the same thing. Estrogen surges during pregnancy drive CBG levels up, meaning total serum cortisol climbs through the trimesters. This does not necessarily reflect higher free cortisol activity, and the magnitude of the estrogen-driven increase is neither predictable nor linear from person to person.2The Journal of Clinical Endocrinology & Metabolism. Salivary Cortisol and Cortisone Measurement Improves ACTH Stimulation Testing in Women Taking Oral Contraceptives For pregnant women or those on oral contraceptives, salivary cortisol is often a better option because it sidesteps the CBG problem entirely.
Why Saliva Samples Have Become So Popular
Salivary cortisol has a straightforward advantage: the cortisol that ends up in your saliva is in equilibrium with the free cortisol in your blood, meaning the unbound, biologically active portion. It is not influenced by changes in CBG.3PubMed. Late-night salivary cortisol as a screening test for Cushing’s syndrome You also do not need a needle stick or a visit to a lab. You chew on a cotton swab or drool into a tube at home, and mail the sample in.
Late-night salivary cortisol has become a frontline screening test for Cushing’s syndrome, a condition caused by prolonged excess cortisol. In healthy people, cortisol drops to its lowest point around 11 p.m. to midnight. People with Cushing’s syndrome lose that nighttime dip, so their late-night values stay high. Studies have consistently shown sensitivity above 90 percent and specificity above 90 percent for this approach.4The Journal of Clinical Endocrinology & Metabolism. Utility of Salivary Cortisol Measurements in Cushing’s Syndrome and Adrenal Insufficiency One study found that setting the midnight salivary cortisol cutoff at a specific threshold achieved 100 percent sensitivity and 96 percent specificity in distinguishing Cushing’s patients from obese controls.5The Journal of Clinical Endocrinology & Metabolism. Midnight Salivary Cortisol for the Initial Diagnosis of Cushing’s Syndrome of Various Causes
Researchers have also explored measuring salivary cortisone alongside cortisol, since cortisone is converted from cortisol by an enzyme in the salivary glands. One study found that combining late-night salivary cortisone with cortisol improved diagnostic accuracy further, with cortisone alone reaching 100 percent specificity at an optimized cutoff.6PubMed Central. Late-night salivary cortisol and cortisone should be the initial screening test for Cushing’s syndrome This dual-marker approach is gaining traction, though not every lab offers cortisone assays yet.
The 24-Hour Urine Collection
A 24-hour urine free cortisol (UFC) test takes a different approach. Instead of capturing a snapshot, it integrates all the cortisol your kidneys filtered over an entire day and night. Because only unbound cortisol passes through the kidney’s filtration system, UFC is not affected by CBG levels the way a blood draw is.7PubMed Central. Urine free cortisol in the diagnosis of Cushing’s syndrome: is it worth doing and, if so, how? That makes it useful in the same populations where total serum cortisol gets muddy, such as people on estrogen therapy.
The practical downside is obvious: you have to carry a jug around for a full day and remember to collect every void. Miss one, and the test underestimates your cortisol. Collect for 26 hours instead of 24, and it overestimates. The test is also vulnerable to how much you drink. Research has shown that drinking about five liters of fluid per day, roughly double a normal intake, raised urinary free cortisol significantly, pushing the majority of samples above the upper normal limit even in healthy people.8PubMed. High fluid intake increases urine free cortisol excretion in normal subjects If you happen to be a heavy water drinker, or if you are collecting during a hot week when you are hydrating aggressively, a mildly elevated urine cortisol result might be misleading.
Clinicians typically ask for two or three 24-hour collections done on separate days to average out that kind of variability. A single elevated result is rarely enough to confirm or exclude a diagnosis.
Timing Is Everything
Cortisol follows a steep daily rhythm. Levels are highest within the first hour after waking and lowest around midnight. A blood draw at 8 a.m. might return a number twice as high as one drawn at 4 p.m. in the same person. This is not a sign of disease; it is the expected pattern. When doctors order a morning cortisol to screen for adrenal insufficiency, they are counting on this peak. When they order a late-night salivary cortisol to screen for Cushing’s, they are counting on the nadir.
There is a distinct phenomenon called the cortisol awakening response (CAR): a sharp spike in cortisol that occurs in the first 30 to 45 minutes after you open your eyes. Researchers who study the CAR use saliva samples collected at waking and then at 15-minute intervals for up to an hour.9PubMed Central. Assessing the daily stability of the cortisol awakening response in a controlled environment Even small delays in sampling can distort results. A delay of more than 15 minutes between actually waking up and collecting the first sample has been shown to reduce the measured spike, potentially making it look blunted when it is not.10Psychoneuroendocrinology. Delays of 5–15 min between awakening and the start of saliva sampling matter in assessment of the cortisol awakening response This is mostly a concern in research settings, but it underscores a broader point: the accuracy of any cortisol test depends heavily on getting the timing right.
The Dexamethasone Suppression Test
Beyond measuring cortisol directly, doctors sometimes test how your body responds to a synthetic steroid called dexamethasone. The overnight dexamethasone suppression test (DST) works like this: you take a low dose of dexamethasone (usually 1 mg) at bedtime, then have your blood cortisol measured the next morning. In a healthy person, dexamethasone signals the brain to dial down cortisol production, so the morning value drops. If cortisol remains elevated, it suggests the feedback loop is not working properly, which is a hallmark of Cushing’s syndrome.
The traditional cutoff for a “normal” suppression was below 5 µg/dL, but research demonstrated that this threshold was too generous and missed some patients whose cortisol suppressed partially but not fully. A study of 190 healthy subjects found that only 5 had morning cortisol above 2 µg/dL after dexamethasone, leading to the recommendation that the normal cutoff should be below about 1.8 to 2 µg/dL.11Steroids. Overnight dexamethasone suppression test: normal responses and the diagnosis of cushing’s syndrome More recent work has suggested that slightly higher thresholds, around 2.1 to 2.3 µg/dL depending on the suspected cause of Cushing’s, may further optimize accuracy in specific clinical scenarios.12PubMed Central. Optimizing Diagnostic Accuracy in Cushing Syndrome Using 1 mg Dexamethasone Suppression Test Cut-Offs
One source of false positives is that some people do not absorb dexamethasone well, or they metabolize it quickly. Their cortisol stays high not because of Cushing’s but because the dexamethasone never reached an adequate blood level. Measuring the dexamethasone concentration in the same blood sample can catch these cases. One study found that simultaneously measuring dexamethasone alongside cortisol reduced false-positive results by about 20 percent.13European Journal of Endocrinology. Simultaneous assay of cortisol and dexamethasone improved diagnostic accuracy of the dexamethasone suppression test
Medications That Interfere With Results
Several common drugs can muddy cortisol testing. Synthetic glucocorticoids like prednisolone and methylprednisolone are structurally similar to cortisol, and some immunoassays used in standard labs cross-react with them. That means the assay reads the drug as if it were cortisol, inflating the result. Prednisolone and 6-methylprednisolone are particularly problematic, with high cross-reactivity and a high likelihood of clinically significant false elevations in patients taking those drugs.14PubMed Central. Cross-reactivity of steroid hormone immunoassays: clinical significance and two-dimensional molecular similarity prediction
This is one reason why the lab method matters. Most routine cortisol measurements use immunoassays, which are fast and cheap but susceptible to cross-reactivity. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is the more precise alternative, able to distinguish cortisol from structurally similar compounds. Research comparing the two approaches in saliva samples has found nonlinear relationships between them, meaning an immunoassay and LC-MS/MS can give meaningfully different numbers from the same sample.15PubMed. Comparison of salivary cortisol as measured by different immunoassays and tandem mass spectrometry If you are on a medication known to cross-react, or if initial test results seem inconsistent with your symptoms, your doctor may request that the lab use mass spectrometry.
Children Need Different Reference Ranges
A pitfall that does not get enough attention is that children’s cortisol levels are naturally lower than adults’. When pediatric blood samples are flagged using adult reference ranges, a striking number come back labeled “low” even though the child’s cortisol is perfectly normal for their age. One analysis found that using adult reference ranges on a standard platform flagged about 35 percent of samples from children aged 1 to 12 as abnormally low, compared to only 7 percent when age-specific pediatric ranges were applied.16The Journal of Clinical Endocrinology & Metabolism. A-335 Pediatric Cortisol Reference Intervals (RI): Impact of Age-Specific RI on the Rate of Abnormal “Low” Cortisol Results On another platform, roughly half of infant samples were flagged low using adult norms. This can trigger unnecessary follow-up testing and parental anxiety. If your child’s cortisol comes back marked low, it is worth asking whether the lab used pediatric-specific reference ranges.
Hair Cortisol for the Long View
Blood, saliva, and urine all capture cortisol over a span of minutes to hours. Hair cortisol works on a completely different timescale. Cortisol is incorporated into the hair shaft as it grows, and since hair grows roughly one centimeter per month, a one-centimeter segment snipped from the scalp can estimate your average cortisol exposure over the past month. A three-centimeter segment reflects the past three months, analogous to the way hemoglobin A1c reflects average blood sugar over a similar period.17PubMed Central. Hair Cortisol Analysis: A Promising Biomarker of HPA Activation in Older Adults
This makes hair cortisol uniquely suited for questions about chronic stress. Where saliva and urine capture real-time levels, hair provides a retrospective window that no other sample type can match.18PubMed. Hair cortisol as a biological marker of chronic stress: current status, future directions and unanswered questions Researchers have used it to study populations where repeated sampling would be impractical: people exposed to natural disasters, caregivers of chronically ill family members, and populations in remote areas without regular lab access. Hair cortisol is not yet part of routine clinical diagnostics for conditions like Cushing’s, but it fills a niche that blood and saliva cannot.
The obvious limitation: you need hair. People who are bald, who shave their heads, or who have very short or chemically treated hair may not have a usable sample. Researchers have begun exploring fingernail cortisol as an alternative keratinized tissue. Cortisol accumulates in nails similarly to hair, and one validation study found that fingernail cortisol correlated with both hair cortisol and cumulative salivary cortisol measured months earlier.19PubMed. A validation study on fingernail cortisol: correlations with one-month cortisol levels estimated by hair and saliva samples Nail cortisol is still very much a research tool, but the concept is the same: a keratinized tissue that archives cortisol over weeks to months, offering a window into chronic exposure that a single-morning blood draw could never provide.20PubMed. Human nail cortisol as a retrospective biomarker of chronic stress: A systematic review
When the Standard Tests Aren’t Enough
For the majority of people, the combination of blood cortisol, salivary cortisol, urine free cortisol, and the dexamethasone suppression test is sufficient to diagnose or rule out cortisol disorders. But in cases where Cushing’s syndrome is confirmed and the question shifts to where the excess hormone is coming from, more invasive testing sometimes enters the picture.
Inferior petrosal sinus sampling (IPSS) is considered the gold standard for distinguishing a pituitary source of excess ACTH from an ectopic source, such as a tumor elsewhere in the body. The procedure involves threading catheters into veins that drain the pituitary gland and measuring ACTH concentrations before and after stimulation. A higher ACTH concentration near the pituitary compared to a peripheral vein confirms a pituitary source.21PubMed Central. Pitfalls in Performing and Interpreting Inferior Petrosal Sinus Sampling: Personal Experience and Literature Review It is not a routine screening test. IPSS is reserved for patients with established Cushing’s whose imaging and biochemical tests point in conflicting directions. Recent work has shown that adjusting ACTH ratios by simultaneously measuring prolactin during IPSS improves diagnostic accuracy, correctly identifying the source in all patients in one series where unadjusted ratios produced three incorrect diagnoses.22PubMed Central. Prolactin-adjusted inferior petrosal sinus sampling: Pituitary and ectopic adrenocorticotropic hormone-dependent Cushing syndrome
Cortisol Testing in Critical Illness
Acute illness complicates cortisol testing in ways that go beyond simple confounders. In critically ill patients, the normal cortisol response may be disrupted by inflammation, changes in blood proteins, and altered metabolism of the hormone. A condition known as critical illness-related corticosteroid insufficiency (CIRCI) describes a situation where the adrenal response is inadequate for the level of physiological stress the body is under. Diagnosing it is tricky because the usual cutoffs for “normal” or “low” cortisol do not apply in the ICU setting the way they do in an outpatient clinic.23PubMed Central. Critical illness-related corticosteroid insufficiency (CIRCI) – an overview of pathogenesis, clinical presentation and management CBG levels often drop during severe illness, which lowers total serum cortisol on paper while free cortisol may actually be elevated. Salivary cortisol and free cortisol measurements are sometimes used in these settings to get around the CBG problem, though there is no universally agreed-upon protocol.
Wearable Sensors and the Future of Cortisol Monitoring
All existing clinical cortisol tests share a limitation: they give you a single data point (or at best a handful of data points over 24 hours). Researchers are working on wearable devices that could monitor cortisol continuously through sweat. Electrochemical biosensors designed for point-of-care use have emerged as a promising technology for real-time cortisol tracking.24PubMed. Recent developments in wearable & non-wearable point-of-care biosensors for cortisol detection
One recently described prototype, called “Stressomic,” integrates sweat extraction with microfluidic channels on a wearable patch. It tracks cortisol alongside epinephrine and norepinephrine, producing a continuous profile of the stress response during physical exertion, psychological stress, and pharmacological challenges.25PubMed Central. Stressomic: A wearable microfluidic biosensor for dynamic profiling of multiple stress hormones in sweat These devices are still in early-stage human studies and are not available clinically, but the direction is clear: the field is moving toward continuous monitoring that could eventually make the once-a-day blood draw look primitive. Advances in nanomaterials and sensor design continue to push these platforms closer to something that could realistically be used outside a research lab.26PubMed. Cortisol as a stress biomarker: analytical perspectives and challenges
For now, the practical challenge is calibration. Sweat cortisol levels do not map neatly onto blood or saliva levels, and sweat rate itself varies with temperature, fitness, and how much skin the sensor covers. Until large validation studies establish clinical reference ranges for sweat cortisol, these devices are likely to find their first real uses in research and wellness tracking rather than in diagnosing medical conditions.