A guide to cortisol and every way of testing it. Information, not medical advice. Read more

Understanding cortisol

Normal cortisol levels: what the ranges mean for each test

A normal cortisol level depends on the test, the time of day and the laboratory. Here are typical ranges for blood, saliva, urine and hair, the thresholds doctors use, and how to convert between units.

There is no single normal cortisol level. What counts as normal depends on the type of test, the time of day the sample was taken and the laboratory that measured it. For a blood test at 8 am, one US reference gives 5 to 25 µg/dL, which is 138 to 690 nmol/L.1 A UK hospital laboratory quotes 133 to 537 nmol/L for samples taken between 6 and 10 am, and below 50 nmol/L at midnight.2 This guide sets out typical values for each kind of test, the thresholds doctors use, and why the range printed on your own report is the one that counts.

Why there is no single normal range

Three things move the goalposts.

  • Time of day. In a healthy reference group, blood cortisol peaked at about 15.5 µg/dL at around 8.30 am and fell below 2 µg/dL around midnight.3 A result means little without the time it was taken. One NHS laboratory states that a random cortisol, taken at a time other than 9 am, cannot be interpreted.4 See the cortisol daily rhythm.
  • The laboratory's method. Five automated immunoassays read 1.08 to 1.36 times higher than a mass spectrometry reference method.5 When one NHS laboratory changed its assay, it warned that results would read about 20 per cent lower.2
  • Hormones and medicines. Oral oestrogen raises the protein that carries cortisol in the blood. In one study, women taking it had 67 per cent higher total blood cortisol.6 Steroid medicines, including inhalers, creams and injections, can lower a morning result.7

So treat every range below as an example. Compare your own result with the range on your report and, if you have symptoms, discuss it with the doctor who ordered the test.

Morning blood cortisol: typical ranges

Blood (serum) cortisol is usually measured in the morning. These published examples each apply only to that laboratory and its assay.

LaboratoryWhen takenRange (nmol/L)Range (µg/dL)
University Hospitals Birmingham, Abbott assay8Before 10 am101 to 5363.7 to 19.4
University Hospitals Birmingham, Abbott assay8After 5 pm79 to 4782.9 to 17.3
Worcestershire Acute Hospitals, Beckman Coulter assay99 am185 to 6246.7 to 22.6
Royal Liverpool26 to 10 am133 to 5374.8 to 19.5
Royal Liverpool2MidnightBelow 50Below 1.8
MedlinePlus, US reference18 am138 to 6905 to 25

The µg/dL figures for the UK laboratories are our own conversion, at 27.59 nmol/L per µg/dL.10 Notice how far the lower limits differ, from 101 to 185 nmol/L, for samples taken at much the same time of morning.89

What doctors look for in a low morning result

For anyone aged 16 or over with suspected adrenal insufficiency, NICE uses a blood cortisol taken between 8 and 9 am, read against decision thresholds rather than a reference range:7

  • Below 150 nmol/L: the person may have adrenal insufficiency and should be referred to a hormone specialist (endocrinologist).
  • 150 to 300 nmol/L: uncertain; consider repeating the test.
  • Above 300 nmol/L: adrenal insufficiency is very unlikely.

These cut-offs apply only to modern immunoassays. NICE also advises stopping oral oestrogen for six weeks before the test, and waiting four weeks after a steroid injection into a muscle or joint.7 NICE's guidance for GPs adds that a result below 100 nmol/L makes adrenal insufficiency highly likely and calls for hospital admission.11

A low or borderline result is followed by a Synacthen test, in which cortisol is measured before and after an injection of synthetic ACTH. With healthy adrenal glands it rises above 500 to 550 nmol/L, and a peak below 500 nmol/L is diagnostic of primary adrenal insufficiency, although exact cut-offs depend on the local assay.11 See the Synacthen test.

Very low cortisol can turn into an adrenal crisis, which is an emergency. If someone with known or suspected adrenal insufficiency suddenly gets worse, with severe dizziness, vomiting, confusion, drowsiness or collapse, call 999.12

Normal saliva cortisol, late at night and in the morning

A morning blood cortisol is not one of the recommended first tests for Cushing's syndrome. Instead, doctors check whether cortisol falls as it should, using late-night saliva, 24-hour urine or a dexamethasone suppression test.13 Saliva measures free cortisol, so it is not affected by the oestrogen effect on blood tests.5

In healthy people with regular sleep, saliva cortisol at bedtime or between 11 pm and midnight is below 4 nmol/L (145 ng/dL).13 One NHS laboratory uses these locally derived ranges:14

Saliva sampleRange (nmol/L)
Late night (11 pm to midnight)Below 3.2
Early morning (7 to 9 am)3.2 to 22.8

Levels around waking are harder to pin down. Saliva cortisol normally rises by 50 to 160 per cent in the first 30 minutes after you wake, an average rise of about 9 nmol/L.15 There are no agreed norms for the absolute level: studies report 4.7 to 18.5 nmol/L on waking and 8.6 to 21.9 nmol/L half an hour later.15 Because results vary from night to night, the Endocrine Society recommends at least two late-night samples.13 See the saliva cortisol test.

Normal 24-hour urine cortisol

A 24-hour urine collection measures the free cortisol you pass over a whole day. One US reference gives normal as below 124 nmol (45 µg) per 24 hours for women and below 165 nmol (60 µg) for men.16 A study of 152 people with Cushing's disease used a range of 30 to 145 nmol per 24 hours, specific to its assay.17

The same study showed why one collection is rarely enough: results varied by about 52 per cent from one collection to the next in the same person.17 Guidelines recommend at least two collections.13 See the urine cortisol test.

The dexamethasone suppression test cut-off

Here, normal means cortisol switching off. You take a low dose (1 mg) of dexamethasone, a synthetic steroid, at night and have blood taken the next morning. In a healthy person the 8 am cortisol is then below 50 nmol/L (1.8 µg/dL).18 A result above that is the positive cut-off used when investigating Cushing's syndrome.13 Oral oestrogen can make the test misleading: false-positive rates reach 50 per cent in women on the combined pill, so oestrogen is stopped for six weeks beforehand where possible.13 See the dexamethasone suppression test.

Normal hair cortisol: why every laboratory's range differs

Hair cortisol is reported in picograms per milligram of hair (pg/mg). Some papers use ng/g, which gives the same number.19 There is no single normal range. Reported values for healthy adults differ more than tenfold depending on the laboratory method:2019

StudyMethodPeopleHealthy values (pg/mg)
Fortuna 202620LC-MS/MS256 healthy adults, Northern Italy0.84 to 8.07 (median 2.41)
Wester 201721ELISA immunoassay174 healthy controlsGeometric mean 8.4
Manenschijn 201322Not stated283 older adultsWomen 21.0, men 26.3
Sauvé 200719Adapted saliva immunoassayNon-obese adults17.7 to 153.2 (median 46.1)

So a hair result can only be read against the range of the laboratory that measured it. What hair offers instead of a universal range is consistency. One sample averages roughly the last three months,23 and the same people's hair cortisol correlated at r = 0.84 from one month to the next, against 0.38 to 0.61 week to week for saliva and 0.59 to 0.68 for 24-hour urine.24 That makes hair the most accurate way to measure your long-term cortisol from a single sample, and makes repeat results from the same laboratory meaningful to compare.

A hair result above the range means more cortisol than most healthy adults over the months tested; below it means less. It does not diagnose or rule out Cushing's syndrome or any other condition, and hair is not part of the diagnostic guidelines.13 See hair cortisol results explained.

Converting between nmol/L and µg/dL

UK and European laboratories report blood and saliva cortisol in nanomoles per litre (nmol/L). US laboratories usually use micrograms per decilitre (µg/dL). To convert µg/dL to nmol/L, multiply by 27.59; to go the other way, divide by 27.59.10 So 18 µg/dL is about 497 nmol/L, usually rounded to 500.10 Our cortisol unit converter does the sum for you. Hair results in pg/mg measure something different, an average over months, and cannot be converted into a blood value.

Questions people ask

What is a normal cortisol level in the morning?

It depends on the laboratory and the exact time. Examples include 133 to 537 nmol/L between 6 and 10 am at one UK hospital and 185 to 624 nmol/L at 9 am at another. A US reference gives 5 to 25 µg/dL (138 to 690 nmol/L) at 8 am. Use the range printed on your own report.

What cortisol level is too low?

For a blood test taken between 8 and 9 am, NICE treats a result below 150 nmol/L as a reason to refer to a hormone specialist. It treats 150 to 300 nmol/L as uncertain and above 300 nmol/L as making adrenal insufficiency very unlikely. A low result is confirmed or ruled out with a Synacthen test.

What cortisol level is too high?

There is no single cut-off from a morning blood test, because cortisol is naturally near its peak then. Doctors check for Cushing's syndrome with tests of whether cortisol falls when it should: late-night saliva, 24-hour urine, and the dexamethasone suppression test, where a morning cortisol above 50 nmol/L (1.8 µg/dL) after dexamethasone is abnormal.

Why is my cortisol result different from last time?

Small differences are expected. Cortisol changes through the day and with stress, and having blood taken can raise it. Oral oestrogen raises total blood cortisol, steroid medicines can lower it, and laboratories sometimes change their assay. A single blood, saliva or urine result reflects the moment or the day it was taken.

What is a normal hair cortisol level?

There is no single range. Reported values for healthy adults differ more than tenfold between laboratory methods, from about 0.8 to 8 pg/mg with mass spectrometry to 17.7 to 153.2 pg/mg with an early immunoassay. Compare your result only with the range on your report, and with your own earlier results from the same laboratory.

How do I convert cortisol from µg/dL to nmol/L?

Multiply by 27.59. To go from nmol/L to µg/dL, divide by 27.59. For example, 18 µg/dL is about 497 nmol/L. Our cortisol unit converter does the sum for you.

Does the contraceptive pill affect cortisol results?

Oral oestrogen, in the combined pill and some HRT, raises the protein that carries cortisol in the blood, so total blood cortisol reads higher. In one study it was 67 per cent higher. Saliva and urine tests measure free cortisol and are not affected in the same way. Hair studies have found at most a small and inconsistent effect.

Sources

  1. MedlinePlus (2025). Cortisol blood test. US National Library of Medicine medlineplus.gov
  2. Royal Liverpool and Broadgreen University Hospitals NHS Trust (undated). Cortisol (serum). Pathology handbook pathlabs.rlbuht.nhs.uk
  3. Debono M et al. (2009). Modified-release hydrocortisone to provide circadian cortisol profiles. J Clin Endocrinol Metab pubmed.ncbi.nlm.nih.gov
  4. Oxford University Hospitals NHS Foundation Trust (2026). Cortisol. Biochemistry test catalogue ouh.nhs.uk
  5. Clarke SA et al. (2024). Current challenges and future directions in the assessment of glucocorticoid status. Endocr Rev pubmed.ncbi.nlm.nih.gov
  6. Qureshi AC et al. (2007). The influence of the route of oestrogen administration on serum levels of cortisol-binding globulin and total cortisol. Clin Endocrinol (Oxf) pubmed.ncbi.nlm.nih.gov
  7. National Institute for Health and Care Excellence (2024). Adrenal insufficiency: identification and management (NG243). NICE guideline nice.org.uk
  8. University Hospitals Birmingham NHS Foundation Trust (2026). Cortisol. Pathology tests database uhb.nhs.uk
  9. Worcestershire Acute Hospitals NHS Trust (undated). Cortisol. Pathology tests A to Z worcsacute.nhs.uk
  10. PubChem, US National Library of Medicine. Hydrocortisone (cortisol): compound summary, molecular weight 362.5 g/mol pubchem.ncbi.nlm.nih.gov
  11. NICE Clinical Knowledge Summaries (2025). Addison's disease. NICE CKS cks.nice.org.uk
  12. NHS (2025). Addison's disease. NHS website nhs.uk
  13. Nieman LK et al. (2008). The diagnosis of Cushing's syndrome: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab pubmed.ncbi.nlm.nih.gov
  14. University Hospitals Birmingham NHS Foundation Trust (2026). Cortisol (saliva). Pathology tests database uhb.nhs.uk
  15. Clow A et al. (2004). The awakening cortisol response: methodological issues and significance. Stress pubmed.ncbi.nlm.nih.gov
  16. MedlinePlus (2025). Cortisol urine test. US National Library of Medicine medlineplus.gov
  17. Petersenn S et al. (2014). High variability in baseline urinary free cortisol values in patients with Cushing's disease. Clin Endocrinol (Oxf) pubmed.ncbi.nlm.nih.gov
  18. MedlinePlus (2025). Dexamethasone suppression test. US National Library of Medicine medlineplus.gov
  19. Sauvé B et al. (2007). Measurement of cortisol in human hair as a biomarker of systemic exposure. Clin Invest Med pubmed.ncbi.nlm.nih.gov
  20. Fortuna F et al. (2026). Establishment of a reference range for hair cortisol using a HPLC-MS/MS method in a Northern Italian population. J Endocrinol Invest pubmed.ncbi.nlm.nih.gov
  21. Wester VL et al. (2017). Scalp hair cortisol for diagnosis of Cushing's syndrome. Eur J Endocrinol pubmed.ncbi.nlm.nih.gov
  22. Manenschijn L et al. (2013). High long-term cortisol levels, measured in scalp hair, are associated with a history of cardiovascular disease. J Clin Endocrinol Metab pubmed.ncbi.nlm.nih.gov
  23. Adav SS & Ng KW (2025). The multifaceted role of hair as a biospecimen: recent advances in precision medicine and forensic science. Exp Mol Med pubmed.ncbi.nlm.nih.gov
  24. Short SJ et al. (2016). Correspondence between hair cortisol concentrations and 30-day integrated daily salivary and weekly urinary cortisol measures. Psychoneuroendocrinology pubmed.ncbi.nlm.nih.gov