Cortisol Reference Values: Morning, Evening and Lab Units

Why there is no single normal range for cortisol

Cortisol normally peaks shortly after waking and declines across the day, so a result is inseparable from its timestamp. A serum cortisol of 8 µg/dL (221 nmol/L) may be unremarkable at 4 pm but deserves context if collected at 8 am during acute illness. In my 15 years of clinical practice, the missing collection time has caused more avoidable alarm than modest numerical abnormalities.

Laboratories build reference intervals from their own population, instrument and calibration materials; they are not universal treatment targets. Kantesti is an AI blood test analyzer that reads cortisol alongside collection time, units, specimen and adjacent adrenal markers rather than colouring one number red. This is particularly useful when people compare reports from two countries or two laboratories.

A single cortisol result does not diagnose Cushing syndrome or adrenal insufficiency. The Endocrine Society advises against using random serum cortisol to screen for Cushing syndrome because the normal circadian swing is too wide (Nieman et al., 2008). For related hormone context, see our guide to cortisol and ACTH patterns.

The clock time belongs beside the result

A cortisol report without a collection time is clinically incomplete for most outpatient questions. Record whether you woke at 5 am or worked overnight, because the biological morning peak follows the sleep-wake cycle more closely than the wall clock in shift workers.

Morning cortisol levels: useful intervals and diagnostic cutoffs

For suspected adrenal insufficiency, an 8-9 am cortisol below 3 µg/dL (83 nmol/L) is strongly concerning when symptoms and ACTH fit; a value above 15 µg/dL (414 nmol/L) often indicates adequate reserve in stable outpatients. Values between 3 and 15 µg/dL are an indeterminate zone, not a diagnosis, and usually prompt an ACTH stimulation test or specialist review.

The old stimulated cortisol target of 18 µg/dL (500 nmol/L) cannot simply be carried into every modern assay. Javorsky and colleagues found that assay-specific stimulated cutoffs near 14-15 µg/dL may prevent false diagnoses of adrenal insufficiency with newer assays (Javorsky et al., 2021). That technical detail changes real lives; I have seen patients unnecessarily started on glucocorticoids after an outdated threshold was applied.

If your report says 7.2 µg/dL at 07:45, do not convert it into a verdict before checking illness, sleep, steroid exposure and albumin. The practical next comparison is usually a properly timed repeat with ACTH sample handling planned in advance, because ACTH is far less stable than cortisol.

Evening cortisol levels should be lower than morning values

At 4-6 pm, a total serum cortisol of 12 µg/dL (331 nmol/L) can sit above a laboratory's expected interval even though that same number would be ordinary at 8 am. This is why comparing an evening draw to a morning screenshot online is unsafe. The direction of change across the day can matter as much as the absolute value.

Late-night testing is used mainly when clinicians suspect cortisol excess, not because every tired person needs a nighttime cortisol check. Late-night salivary cortisol is normally very low near bedtime; each laboratory sets its own upper limit, often around 0.09-0.15 µg/dL (2.5-4.1 nmol/L) depending on assay and collection device.

An overnight worker who sleeps from 09:00 to 16:00 may have a shifted cortisol rhythm, making conventional clock-time intervals less dependable. In these cases I ask for sleep timing, recent travel and light exposure before interpreting results; our article on jet lag and test timing explains why this history changes the plan. Kantesti is an AI blood test interpretation platform that preserves this timing context when comparing hormone reports.

Cortisol lab units: µg/dL, nmol/L and ng/mL

The conversion formula is simple: µg/dL multiplied by 27.59 equals nmol/L, while nmol/L divided by 27.59 equals µg/dL. Thus 10 µg/dL is 276 nmol/L, and 500 nmol/L is 18.1 µg/dL. A report using ng/mL shows 10 µg/dL as 100 ng/mL, which looks ten times larger but is the same concentration.

Do not apply that formula blindly to urinary free cortisol reporting in µg per 24 hours or nmol per 24 hours. Those units describe total hormone excreted over a timed collection, not a concentration at one moment. A collection of 60 µg/24 hours may be within one laboratory's limit and above another's, depending on methodology and local validation.

A surprisingly common error is confusing µg/dL with µg/L; they differ by a factor of 10. Dr. Thomas Klein recommends saving the original PDF rather than manually copying values into notes, especially when reviewing a hormone panel with other DHEA-S clues.

Serum, saliva and urine cortisol measure different things

Total serum cortisol includes cortisol bound to corticosteroid-binding globulin, albumin and the small free fraction. High estrogen states raise corticosteroid-binding globulin and can raise total cortisol without true cortisol excess. Low albumin or markedly low binding globulin can do the reverse, leaving a total serum value deceptively low.

Salivary cortisol is useful near bedtime because only free hormone diffuses into saliva, but contamination matters. Brushing teeth, oral bleeding, steroid creams transferred from fingers, tobacco and eating close to collection can distort a low-level sample. Our detailed comparison of saliva and urine cortisol tests covers when each specimen is sensible.

Twenty-four-hour urinary free cortisol averages secretion across a day and is generally interpreted against the collection's upper limit of normal, often around 50 µg/24 hours (138 nmol/24 hours) in a given laboratory. Kantesti AI interprets cortisol results by separating total serum, late-night salivary and urinary free cortisol rather than merging them into one score.

Why test methods produce different cortisol ranges

Automated immunoassays are fast and widely available, but prednisolone and some steroid metabolites may interfere depending on the manufacturer. LC-MS/MS separates molecules by mass and is especially helpful when results conflict with symptoms or when exogenous steroid exposure is plausible. Neither method is inherently “wrong”; the lab-specific interval must match the method used.

Assay bias becomes most visible at low cortisol concentrations, exactly where late-night salivary cortisol and adrenal testing decisions can be sensitive. A 2 µg/dL difference around a stimulated cutoff may change whether someone is labelled deficient. Bornstein and colleagues advise assay-aware interpretation in suspected primary adrenal insufficiency rather than treating one historical cutoff as universal (Bornstein et al., 2016).

Kantesti AI flags method changes when they are visible on a report and avoids declaring a trend when an immunoassay has been replaced by LC-MS/MS. Readers can review our clinical validation approach to see why source ranges and collection metadata matter as much as optical character recognition.

Sleep, stress and illness can change a cortisol result quickly

Cortisol begins to rise about 2-3 hours before usual waking, then often increases a further 38-75% during the first 30-45 minutes after awakening. That cortisol awakening response is normal physiology, which is why an 06:30 draw after a 04:30 awakening cannot be interpreted exactly like an 08:00 draw after an undisturbed night.

Fever, major dental procedures, a marathon, panic, severe calorie restriction and hospitalization can all increase serum cortisol. Conversely, severe critical illness can alter binding proteins so total cortisol and free cortisol diverge. I see this pattern often after poor sleep; our practical guide to labs after sleepless nights helps distinguish a repeat-worthy result from an urgent one.

For a planned baseline test, keep wake time as usual, avoid unusually hard training for 24 hours and sit quietly for 15-30 minutes before collection if the laboratory permits. Do not try to “lower cortisol” by skipping prescribed medicine or fasting longer than instructed; that can create a less useful, not more accurate, result.

How to compare cortisol results without creating a false trend

Start with four fields: exact collection time, reported unit, specimen type and reference interval printed beside the result. Add wake time, recent steroid doses, pregnancy status, acute illness and whether the laboratory changed. This small audit often explains an apparent rise from 9 to 16 µg/dL without any new disease.

If repeats are being used to evaluate an endocrine problem, try to collect at the same clock time within 30-60 minutes and use the same lab. A morning serum result from one laboratory and a late-night saliva result from another should be treated as complementary tests, not a rising or falling series. Our side-by-side result comparison guide gives a patient-friendly record format.

Kantesti is an AI-powered blood test analysis tool that compares cortisol only after checking whether the results share a meaningful clinical context. Its trend view cannot replace dynamic endocrine testing, but it can surface the question a clinician needs answered: “Were these actually comparable samples?”

Pregnancy and estrogen can raise total cortisol without Cushing syndrome

During pregnancy, total serum cortisol can reach roughly 2-3 times non-pregnant values by the third trimester, and the daily rhythm is also less sharply defined. Standard non-pregnant serum reference intervals are therefore poor tools for screening cortisol excess in pregnancy. Clinicians usually choose pregnancy-aware specialist testing when symptoms are compelling.

Combined oral contraceptives and oral estrogen can increase binding globulin over several weeks; transdermal estrogen usually has a smaller hepatic binding effect. In contrast, nephrotic protein loss, severe liver disease and critical illness may lower binding proteins and reduce total cortisol. The result is a biochemical illusion unless the context is visible.

A person taking an estrogen-containing pill with a morning cortisol of 28 µg/dL (773 nmol/L) does not automatically have Cushing syndrome. I would ask about weight change, bruising, new diabetes, blood pressure and physical findings before choosing a test. Hormone timing also matters in men, as discussed in our morning hormone testing guide.

Steroids and common medicines can distort cortisol testing

Hydrocortisone is chemically identical to cortisol and will generally appear as cortisol in serum testing. Prednisolone may cross-react in some immunoassays, while dexamethasone usually has little direct assay cross-reactivity but can suppress ACTH and endogenous cortisol. The exact washout plan depends on the indication, dose and clinical risk, so it belongs with the prescriber or endocrinologist.

Inhaled steroids, potent skin preparations, joint injections and “natural” supplements can cause clinically relevant suppression, particularly with repeated or high-dose exposure. Enzyme-inducing drugs such as carbamazepine, phenytoin and rifampicin can speed cortisol metabolism and complicate replacement therapy. Ritonavir can have the opposite practical problem by increasing exposure to some inhaled steroids.

The term “adrenal fatigue” is not an accepted endocrine diagnosis, and supplements marketed for it can contain undisclosed steroids or stimulants. Before purchasing products to alter a laboratory number, read our evidence-based review of adrenal supplements.

When a low cortisol result needs urgent medical review

A morning cortisol below 3 µg/dL (83 nmol/L) with compatible symptoms should prompt urgent clinician contact, especially after recent steroid withdrawal, pituitary disease, adrenal surgery or severe infection. In a suspected adrenal crisis, clinicians draw cortisol and ACTH if this does not delay care, then administer stress-dose glucocorticoid treatment. Waiting for a repeat home sample is unsafe.

The pattern matters: low cortisol plus high ACTH, low sodium and high potassium supports primary adrenal insufficiency, while low or inappropriately normal ACTH suggests a pituitary or hypothalamic cause. Potassium may remain normal in secondary adrenal insufficiency because aldosterone is often preserved. That distinction guides both imaging and replacement treatment.

If symptoms are milder but persistent, arrange timely review rather than assuming fatigue alone proves cortisol deficiency. Our article on Addison disease clues explains the symptom-and-lab pattern that merits active follow-up.

A high cortisol result rarely confirms Cushing syndrome by itself

The Endocrine Society recommends late-night salivary cortisol, 24-hour urinary free cortisol or a 1 mg overnight dexamethasone suppression test as initial options when clinical suspicion is genuine (Nieman et al., 2008). Typical concerning features include progressive central weight gain, wide purple striae, easy bruising, proximal muscle weakness, new diabetes and difficult hypertension. Fatigue, anxiety and a high single morning cortisol are not enough.

For urinary free cortisol, laboratories often consider values above the assay-specific upper limit on at least two complete collections more persuasive than one elevated collection. A result more than 3 times the upper limit of normal is harder to dismiss, but alcohol excess, uncontrolled diabetes, major depression and acute illness can still create a pseudo-Cushing pattern.

A 52-year-old runner with a cortisol of 27 µg/dL after an overnight emergency department visit needs a different conversation from someone with repeated abnormal late-night saliva samples and worsening diabetes. For the broader differential, see causes of high cortisol.

Dynamic cortisol tests answer questions a baseline value cannot

In a standard ACTH stimulation test, cortisol is sampled before and usually 30 and 60 minutes after synthetic ACTH administration. With newer assays, a peak around 14-15 µg/dL may be adequate, whereas some laboratories retain a cutoff near 18 µg/dL; use the test laboratory's protocol. Baseline ACTH should ideally be collected before treatment when safe.

A 1 mg overnight dexamethasone suppression test normally suppresses next-morning serum cortisol to below 1.8 µg/dL (50 nmol/L) in many protocols. Estrogen, poor dexamethasone absorption, enzyme-inducing drugs and missed doses can yield false-positive results. This is one area where a specialist's pre-test instructions are far more valuable than improvisation.

A 24-hour urine collection needs every void from the start time through the finish time, stored exactly as directed. Under-collection is common and can falsely reassure; review our 24-hour urine collection checklist before repeating a borderline result.

Seven mistakes people make when comparing cortisol reports

First, do not compare a 17:00 serum cortisol with an 08:00 interval. Second, do not convert a late-night saliva value into an “equivalent” serum value. Third, do not assume a change from 250 to 300 nmol/L is clinically meaningful if sleep, stress or assay changed between draws.

Fourth, do not omit steroid exposure because a cream, inhaler or injection seems unrelated. Fifth, do not use a single cortisol test to explain nonspecific tiredness without checking anemia, thyroid disease, sleep, mood, glucose and medication effects. Sixth, do not self-treat a low value with glandular extracts or unregulated corticosteroids.

Seventh, do not confuse “out of range” with dangerous. A result can be statistically unusual yet medically minor, while a value inside range can still be concerning if it is inappropriately low during severe illness. Our guide to out-of-range flags is a useful reset when a portal label feels alarming.

Research sources, AI context and the next safe step

Dr. Thomas Klein’s rule in clinic is simple: retain the original report, write down wake time and all steroid exposures, then bring the complete pattern to the appointment. Kantesti is an AI lab test interpretation service that organizes cortisol with ACTH, sodium, potassium, glucose and prior results, while leaving diagnosis and treatment decisions to licensed clinicians. This prevents a unit mismatch from becoming a frightening story.

Our medical team reviews clinical standards and safety boundaries through the Medical Advisory Board, and the organization’s remit is described on our About Us page. If a report is scanned from a photo, confirm that the extracted unit, decimal point and collection time match the original before relying on any interpretation.

Related Kantesti publications provide methodology context for laboratory interpretation: Kantesti LTD. (2026). BUN/Creatinine Ratio Explained: Kidney Function Test Guide. Zenodo. https://doi.org/10.5281/zenodo.18207872; ResearchGate record and Academia.edu record. Kantesti LTD. (2026). Urobilinogen in Urine Test: Complete Urinalysis Guide 2026. Zenodo. https://doi.org/10.5281/zenodo.18226379; ResearchGate record and Academia.edu record.

Frequently Asked Questions