Cortisol is a hormone released by the adrenal glands on a daily rhythm, and it is tested because that rhythm tracks how the body handles fuel, fluid balance, inflammation and the sleep and wake cycle. Cortisol is usually highest in the hour after waking, falls through the afternoon and reaches its lowest point late at night, so a single number only means something when the collection time is known alongside it. Blood, saliva, urine and hair all carry cortisol, and each captures a different slice of that rhythm. The biomarker page covers what the marker is, how laboratories measure it, its ranges and how a result is read. Everything below is about collecting the sample yourself, outside a clinic, and what each sample type can honestly tell you.
Yes, within limits, and with one important correction to the phrase itself. Saliva, hair and mailed blood or urine kits let you collect a cortisol sample at home, but the sample is analysed by a laboratory, not by you. No consumer device gives you a cortisol number on the spot the way a glucose meter gives a blood sugar reading. What you are doing at home is collecting biological material under controlled conditions and posting it to a lab that runs the assay.
That distinction matters, because almost all of the error in at-home cortisol testing happens before the sample reaches the laboratory. Salivary cortisol is widely used as a non-invasive measure of free, biologically available cortisol in stress research, and its value depends on standardised sampling and careful handling 1. A sample taken forty minutes late, contaminated with blood from vigorous brushing, or left warm in a postbox over a long weekend is not a cortisol result. It is noise with a number attached to it.
The second limit is scope. A home sample can show whether your daily curve has the expected shape, whether the late-night value sits higher than expected, or what average exposure over recent months looks like. It cannot tell you why the number came out the way it did. Working out why is a clinical process that normally involves repeat testing, a review of your medicines and medical history, and in some cases supervised testing that cannot be done at home at all.
Four practical options exist, and they answer genuinely different questions. Choosing by convenience rather than by question is the most common mistake people make.
| Sample | What it reflects | Window covered | How it is collected | Main limitation |
|---|---|---|---|---|
| Saliva, single timed sample | Free cortisol at one moment | Minutes | Passive drool or an absorbent swab into a tube | Close to meaningless without the exact clock time |
| Saliva, series across one day | The shape of the daily curve | One waking day | Three to five tubes at set times | Every tube has to be on time, or the curve distorts |
| Late-night saliva | Whether the expected overnight low point is present | A single evening | One tube, often repeated on a second night | Shift work, late meals and broken sleep change it |
| Hair | Average cortisol laid down as the hair grew | Weeks to months | A small lock cut close to the scalp at the back of the head | Growth rate, washing and chemical processing affect it |
| Dried blood spot | Total cortisol in blood at one moment | Minutes | Finger prick onto a collection card | Needs a well filled spot and an accurate recorded time |
| Dried urine, timed | Cortisol and its breakdown products excreted over set parts of the day | Hours | Urine onto collection strips at set times | Fluid intake and bladder habits shift the picture |
Saliva answers questions about the rhythm. Hair answers questions about the months behind you. Dried blood answers a narrow question about one moment and is closest to what a clinic draw measures. Dried urine sits in between, summarising output over blocks of hours rather than instants.
Saliva reflects the free, unbound fraction of cortisol, while a standard blood test reports total cortisol, most of which is travelling bound to carrier proteins. That is the practical reason saliva is used so widely outside clinics. It needs no needle, it can be collected at the exact minute the protocol calls for, and it tracks the fraction of the hormone that is available to tissues. Salivary cortisol has become a standard non-invasive marker of free cortisol in stress research for exactly these reasons 1.
The trade-off is sensitivity to collection conditions. Saliva volumes are small, cortisol concentrations in saliva are far lower than in blood, and anything that changes the composition of what is in your mouth can change the number. Blood from irritated gums carries total cortisol into the sample. Food residue, acidic drinks and recent caffeine alter flow rate and pH. Published recommendations for salivary sampling cover these issues directly and set out how to avoid eating, drinking and oral bleeding before collection 3.
There is also a units problem that catches people out. Saliva results are typically reported in nmol/L or ng/mL, and they are not comparable with a serum cortisol figure in µg/dL. A saliva result is read against the laboratory's own saliva reference intervals for that time of day, not against a morning blood range.
Because cortisol is a rhythm, not a level. Studies indicate that salivary cortisol follows a clear diurnal profile in healthy volunteers, rising sharply around waking and declining across the day 2. A sample taken at 7am and a sample taken at 7pm can both be entirely expected while looking nothing like each other. Without the clock time, the number has no interpretation.
The first hour after waking is the most sensitive window of all. Research suggests that the cortisol awakening response, the rise over roughly the first thirty to forty-five minutes after waking, is a distinct feature of the rhythm that is highly sensitive to how accurately the sampling times are kept 1. If a kit asks for a sample immediately on waking and another thirty minutes later, dozing for an extra fifteen minutes changes what you have measured.
Three timing rules follow from that.
Follow the kit instructions first, and then apply the standard sampling precautions that published recommendations for salivary cortisol describe 3. The steps below are the common core of almost every protocol.
If you are also tracking a second salivary marker, published recommendations exist for collecting cortisol and alpha-amylase from the same sample, including which collection devices suit both 3.
A late-night saliva sample asks one specific question, which is whether cortisol falls to its expected overnight low point. In a typical rhythm the lowest values of the day occur around the hours before and after midnight, so a value that stays elevated then is informative in a way a morning value is not. Late-night salivary cortisol is used clinically as a screening test for Cushing's syndrome, where the normal overnight fall is lost 4.
This is also the at-home test most often ordered by a clinician rather than bought off the shelf, and it is usually repeated on separate nights, because a single evening is easy to disturb. Alcohol, a late heavy meal, an argument, a late gym session or a child waking you up all sit inside the measurement.
Shift work is the biggest confounder. In people who regularly sleep during the day, the overnight low point does not occur at the standard clock time, and a late-night sample collected at midnight may be sampling the wrong part of their rhythm entirely. Tell whoever ordered the test about your sleep schedule before you collect, not after the result comes back. An unexpectedly high late-night value is not a diagnosis of anything by itself. It is a finding worth discussing with a clinician, who decides whether and how to confirm it.
Hair measures accumulated exposure rather than a moment. As hair grows, cortisol is incorporated into the shaft, so a length of hair cut close to the scalp carries a record of the preceding weeks to months. Hair cortisol concentration has been examined as a marker of long-term cortisol exposure, including in the evaluation of Cushing syndrome 5. That makes it the only at-home option that is indifferent to what time you collect it or whether you slept badly last night.
Collection is simple in principle. A small lock, usually from the posterior vertex at the back of the crown, is cut as close to the scalp as possible, kept in order so the scalp end is identifiable, and taped to a card or sealed in foil. The laboratory then segments it. Hair cortisol analysis commonly assumes an average scalp hair growth of roughly one centimetre a month, so the three centimetres nearest the scalp are taken to represent approximately the three most recent months 6.
Several things complicate the result, and the methodological literature on hair cortisol analysis sets out the main ones, including washing frequency, chemical processing of the hair, hair growth rate and sample handling 6. Very short hair, extensive bleaching or dyeing, and hair products that sit heavily on the shaft can all limit what the laboratory can report. Hair also cannot answer any question about rhythm. It will never tell you whether your overnight low point is intact, because it averages that information away.
Both take a sample that would normally be collected in a clinic and dry it onto a card or strip so it can travel by post. For a dried blood spot, you lance a fingertip, let drops fall onto marked circles on a filter card, let the card air dry completely, then seal and send it. Dried blood spot cortisol is a total cortisol measurement, so it sits closer to a clinic blood draw than saliva does, and the same timing logic applies. A morning sample is usually requested within a defined window after waking.
Dried urine kits ask you to wet a collection strip at several set points across a day, typically on waking, mid-morning, late afternoon and before bed, then dry each strip and post them together. These panels report cortisol along with its breakdown products, which gives a picture of output over blocks of hours rather than snapshots.
Two practical cautions apply to both. First, filling the collection medium properly matters, because a thin, partial blood spot or a barely damp strip can be rejected or give an unreliable reading. Second, methods are not interchangeable. Cortisol assays differ in how they perform across serum, urine and saliva, and in how much they cross-react with related steroids, which affects how comparable results are between methods and laboratories 7. A number from one provider should not be lined up against a number from another provider as though they were the same measurement.
Each sample type is compared against reference intervals built for that sample type, that assay and that collection time. There is no single cortisol range that covers all of them. The figures below are for morning serum cortisol, the measurement most people have seen on a standard blood panel.
| Measure | Typical lab reference range | FitScript optimal range |
|---|---|---|
| Cortisol, morning serum | 6 to 23 µg/dL | 10 to 18 µg/dL |
Typical lab reference ranges vary by laboratory, age and sex, and by the exact time the sample was taken. FitScript's optimal range is an optimization target used alongside the standard range, not a diagnostic cut-off, and only a clinician interprets a result in the context of the person it came from.
Saliva results are reported in different units again, usually nmol/L or ng/mL, with separate intervals for morning and late-night samples. Hair results are reported in pg/mg and have no equivalent in the serum numbers above. If a home kit returns a number with no stated units, no reference interval and no record of collection time, it cannot be read meaningfully, and it is reasonable to ask the provider which method they used. FitScript's membership includes 150+ biomarkers tested per year with interpretation by Atlas, FitScript's AI health advisor, which is the kind of context a standalone number lacks.
A single out of range value reflects the conditions of that sample at least as often as it reflects your physiology, which is why no cortisol result is read on its own. The observation, what it may reflect and what to do next are three separate steps.
A higher than expected value may reflect acute stress in the minutes before collection, illness, pain, intense exercise, a disturbed night, pregnancy, or a genuine shift in the rhythm. A late-night value that stays up is the pattern with the most specific meaning, because the overnight low point is normally reliable, and loss of that fall is why late-night salivary cortisol is used as a screening test 4. Even then, screening tests are repeated and confirmed before anyone draws a conclusion.
A lower than expected value may reflect a sample taken later in the day than intended, an under-filled tube, degradation during a slow postal journey, or an assay reading low for that sample type. It can also reflect something clinically relevant, which is why a persistently low morning result, particularly alongside fatigue, light-headedness on standing, weight loss, salt craving or low blood pressure, is worth discussing with a clinician promptly rather than retested casually at home.
In both directions the next step is the same. Note the time, the conditions and anything unusual about that day, keep the report, and take it to a clinician who can decide whether to confirm it, repeat it or test something else entirely. A suboptimal number is a prompt for a conversation, never a conclusion.
Plenty, and most of it has nothing to do with how stressed you feel. Sleep timing and sleep debt move the whole curve, because the rhythm is anchored to your sleep and wake cycle rather than to the clock. Shift work can invert it. Acute illness, injury, surgery and pain raise cortisol. Intense exercise raises it for hours afterwards, so training in the morning before a sample changes what you capture.
Pregnancy raises total cortisol substantially through changes in carrier proteins, and several prescription medicines change either cortisol production or the way an assay reads it. This is why a clinician reviews your full medication list before interpreting any cortisol result, and why no medicine should be started, stopped or adjusted on the strength of a home test.
Collection factors sit in the same category of real effects. Blood in saliva from oral bleeding, food and drink in the mouth, and the choice of collection device all influence salivary measurements, which is why published sampling recommendations address them specifically 3. For hair, washing frequency, chemical processing and growth rate are recognised methodological considerations 6. And across all sample types, differences between assay methods influence what number comes back for the same biological state 7.
Cortisol is rarely informative alone, and the useful companions depend on the question being asked. For rhythm and stress questions, other salivary markers are sometimes collected from the same sample.
For the broader metabolic picture, a clinician may want fasting glucose and HbA1c, since cortisol influences glucose handling, along with a lipid panel, thyroid markers, DHEA-S, and sex hormones where relevant. Sodium and potassium, blood pressure and a morning paired cortisol and ACTH measurement belong firmly in clinician-ordered territory, as do any dynamic tests, which involve timed sampling after a controlled stimulus and cannot be done at home.
The practical point is that a home cortisol kit tells you about one axis on one day. A panel run from a single blood draw, interpreted next to your symptoms, sleep and training load, answers far more. If a home result is the thing that prompts a fuller panel, it has done its job.
Because home sampling is a screening step, and screening findings get confirmed before they mean anything. The methods differ enough to matter. Cortisol assays vary in performance across serum, urine and saliva and in cross-reactivity with related steroids, which limits how directly results compare between methods and laboratories 7. Two kits can give different numbers from the same person on the same day without either being faulty.
Confirmation also needs context a kit cannot see. A clinician reviews your medicines, sleep pattern, recent illness, pregnancy status, training and the exact collection conditions before deciding whether a value is likely real. They may repeat the same test, order a serum or 24 hour urine measurement, or arrange supervised testing. Where a prescription therapy eventually becomes part of the conversation, a licensed provider evaluates whether it is appropriate for the individual, and that decision is never made from a home sample.
If a clinician does order a blood draw as the confirmation step, how you prepare for it matters as much as it did for the saliva kit. Morning timing, fasting instructions and keeping the appointment calm and unhurried all influence what the draw captures.
Any result that is clearly outside the stated range, and any result that arrives alongside symptoms, is worth taking to a clinician rather than repeating at home. That includes unexplained weight change, new purple stretch marks, easy bruising, muscle weakness around the hips and shoulders, rising blood pressure, new or worsening blood sugar problems, persistent severe fatigue, light-headedness on standing, salt craving, or periods that have stopped without explanation.
Symptoms with a normal result still deserve attention. A single timed sample can miss a pattern that repeat testing or a different sample type would show, so a normal number is not a reason to drop a real symptom. Equally, an unexpected number on its own is not a reason to assume the worst, and it is never a reason to ignore an abnormal result either. Bring the report, the collection times and the collection conditions to the appointment, since the context is what makes the number readable.
This article is educational and is not medical advice. Discuss your results, and any treatment decision that follows from them, with a clinician or licensed provider who can review your full history.
Related reading on this site. If a clinician orders a blood draw as the confirmation step, preparing for a cortisol blood test covers what to do on the morning of the test. For other markers people track between appointments, see how to test blood sugar at home and how to test your blood iron at home.
You can collect the sample at home, but a laboratory still runs the analysis. Saliva, hair, dried blood spots and dried urine strips are all collected by you and posted off. There is no consumer device that displays a cortisol number on the spot the way a glucose meter shows blood sugar, so what you are buying is a collection kit plus lab processing.
Whatever time the kit specifies, and the exact clock time has to be recorded. Cortisol follows a daily rhythm, so a morning sample and an evening sample answer different questions. Morning samples are usually anchored to when you woke up rather than to a fixed hour, and late-night samples are collected around the hours before midnight. If you are late, write down the real time rather than the planned one.
Not a full overnight fast, but most saliva protocols ask you to avoid food and drink for about thirty minutes beforehand, including water in some kits, plus no caffeine, alcohol or nicotine in the collection window. Do not brush or floss in the hour before, because bleeding gums can put blood into the sample. Follow the specific instructions that came with your kit.
No. Saliva reflects the free, unbound portion of cortisol, while a standard blood test reports total cortisol, most of which is bound to carrier proteins. They are reported in different units and read against different reference intervals, so a saliva number cannot be compared with a serum number in µg/dL. Both are useful, but they answer slightly different questions.
Under-filling the tube, collecting at the wrong time, blood in the mouth from brushing, eating or drinking too close to collection, and slow or warm transit in the post. Using a different collection device from the one supplied also matters, because the laboratory's reference intervals assume one method. Labelling every tube with the real date and time prevents the most common avoidable problem.
A small lock cut close to the scalp, usually from the back of the crown, with the scalp end kept identifiable. Laboratories segment the hair by length, with the portion nearest the scalp representing the most recent period. Very short hair, heavy bleaching or dyeing, and some styling products can limit what the lab can report, so check the kit requirements before cutting.
No. A home result is a screening-level observation about one sample on one day, or an average over recent months in the case of hair. It cannot identify a cause, and no condition is diagnosed from a single value. A clinician reviews the result alongside your symptoms, sleep, medicines and history, and decides whether confirmation testing is needed.
Keep the report, note the exact collection time and anything unusual about that day, and take it to a clinician rather than repeating the kit on your own. A value that is well outside the stated range, or one that arrives with symptoms such as persistent severe fatigue, light-headedness on standing, unexplained weight change or rising blood pressure, is worth discussing with a clinician promptly.
That depends on why it was tested in the first place, and the clinician who ordered or reviewed it decides. Retesting a saliva curve a day later rarely adds much, because day to day variation is normal. Repeat testing is more useful after something has genuinely changed, such as sleep schedule, workload or a clinical plan, and it should be done with the same method for comparability.