How to Check Your Cortisol Levels (And How to Actually Read the Results)

You suspect your stress hormones are off. Maybe you're wired-but-tired, waking at 3 a.m., gaining weight around your middle, or running on fumes by mid-afternoon. So you ask your doctor to "check your cortisol," you get a single morning blood draw, and you're told it's normal. Yet you still feel terrible.

Here's what almost no one explains: cortisol isn't a number you check once. It's a rhythm — a wave that's supposed to rise sharply after you wake and fall to almost nothing by bedtime. A single measurement is one dot on a graph that only makes sense as a curve. Miss the timing, use the wrong sample type, or read the result without context, and you can be told you're "fine" while your entire stress-hormone system is dysregulated.

And cortisol testing is genuinely confusing because there isn't one "cortisol test" — there are at least seven legitimate methods, measuring different things (a snapshot, a rhythm, a daily total, a months-long average), each with its own right time, sample type, and pitfalls. Order the wrong one for your question and even a technically flawless result won't tell you what you need to know.

This guide walks through every legitimate way to check your cortisol levels — blood, saliva, urine, and hair — when each one is right, and, most importantly, how to actually read the results so you understand what your body is telling you. By the end, you'll know which test answers your specific question, how to collect it so the result is trustworthy, and how to spot the difference between a genuinely normal result and a "normal" that's hiding a broken rhythm.

Why Cortisol Is a Rhythm, Not a Single Number

Cortisol is governed by your HPA (hypothalamic-pituitary-adrenal) axis and follows a strong circadian pattern. It surges 50–75% in the first 30–45 minutes after waking — a spike called the cortisol awakening response (CAR) — then declines steadily across the day to reach its lowest point around midnight. A healthy curve is high in the morning and near-flat at night.

Most dysregulation shows up as a distorted curve, not simply "too high" or "too low." Common patterns include a blunted morning peak (the classic "burned-out" or fatigued profile), an elevated evening/night value that keeps you wired at bedtime, a flattened overall slope, or a curve that's the right shape but shifted too high or too low across the board. Each of those patterns implies a different underlying story and a different next step — which is precisely the information a single value throws away. Because a random single blood draw at 10 a.m. can't distinguish any of these from normal, timing and sample type are everything.

That's also why the shape of your curve carries real information: research links diurnal cortisol parameters — the size of the awakening response and the steepness of the daily decline — to how your body reacts to and recovers from stress (Diurnal cortisol systematic review 2024). You're not just measuring a level; you're measuring a system's regulation.

There's one more wrinkle that trips people up: most blood tests measure total cortisol, the vast majority of which is bound to proteins and biologically inactive. Only the small "free" fraction actually reaches your tissues and drives symptoms. Saliva and urine tests measure that free, active fraction directly — another reason the sample type you choose changes what you're really seeing. Keep that distinction in mind as we walk through each method below.

1. Morning Blood Cortisol (Serum)

A morning serum cortisol, drawn between roughly 7 and 9 a.m., is the standard starting test your conventional doctor will usually reach for first. Because cortisol peaks in the morning, this captures your presumed high point and is the go-to when clinicians suspect either very high cortisol (Cushing's syndrome) or dangerously low cortisol (adrenal insufficiency or Addison's disease). It's fast, cheap, and widely available, which is why it's the default — but "default" and "right for your question" aren't always the same thing.

What it's good for: confirming a grossly high or low level and screening for adrenal disease. If a clinician suspects adrenal insufficiency, this is often paired with an ACTH stimulation test to see whether your adrenal glands can respond when prompted. Its limitation: a single dot. A "normal" morning value tells you nothing about whether your cortisol falls appropriately at night, which is where most subtle dysregulation hides. It also measures total cortisol, so a high binding-protein level (from estrogen, pregnancy, or oral contraceptives) can nudge the number without reflecting how much active hormone reaches your tissues. It's a necessary test in the right context, but rarely the whole story.

2. Late-Night Salivary Cortisol

This is one of the most clinically validated tests for detecting excess cortisol — and it's clever in its logic. Saliva collected around 11 p.m.–midnight should show cortisol at its daily low, because a healthy curve has bottomed out by then. In early Cushing's syndrome, the very first thing to break is often that night-time trough: the level stays stubbornly elevated when it should be near zero. So sampling at exactly the moment the curve is supposed to be flat gives you an unusually sensitive early-warning signal, sometimes flagging a problem before daytime numbers look abnormal at all. The Endocrine Society's clinical practice guideline endorses late-night salivary cortisol as a first-line screen precisely because it targets the moment the curve should be flat (Nieman 2008).

Salivary testing measures free, biologically active cortisol (not the protein-bound fraction), it's painless, and you can collect it at home at the exact right moment — a real advantage over needing a lab draw at midnight, which itself can spike cortisol just from the stress of the needle. For accuracy, you typically avoid eating, drinking, or brushing your teeth for 30 minutes before the sample, since food residue and blood from your gums can contaminate the result. A single elevated late-night value usually prompts a repeat or a confirmatory second test before any diagnosis.

3. The Four-Point Salivary Cortisol Curve

This is the test that most closely answers the question "is my cortisol rhythm off?" You collect saliva at four points — waking, midday, evening, and bedtime — to map the entire daily curve rather than a single instant. It's the closest thing to watching the whole wave rise and fall instead of freezing one frame of it, and it's usually a simple at-home kit: four small tubes, four labeled times, mailed back to the lab.

The four-point curve is where you actually see the burned-out flat-morning pattern, the wired-at-night elevation, or the sagging afternoon crash that a single blood test misses entirely. Reading it is intuitive once you know the ideal shape: a strong morning value, a steady step-down through midday and evening, and a low bedtime value. Deviations point to specific problems — a low morning number suggests a blunted awakening response and often tracks with fatigue and difficulty getting going, while a high evening or bedtime number is the signature of the "tired but wired" insomnia so many people describe.

It's the workhorse of functional and root-cause testing for fatigue, sleep, and stress complaints, because it turns "your cortisol is normal" into a picture of when and how your rhythm deviates. The one caveat: because saliva reflects a single day, a stressful or poorly-slept test day can distort results, so it's best collected on a typical, representative day.

4. Cortisol Awakening Response (CAR)

A more refined version of morning testing, the CAR specifically measures the rise from the moment you wake to about 30–45 minutes later — not just the level, but how much your body ramps up as it comes online for the day. Think of it as a stress test of your HPA axis's ability to mount its most important daily surge. A healthy CAR is a robust spike; a blunted CAR has been associated with burnout, chronic fatigue, and depression, while an exaggerated CAR can reflect anticipatory stress about the day ahead. It's one of the most researched cortisol markers precisely because it's so tightly linked to psychological load.

The catch is that CAR is highly sensitive to how you collect it. Samples must be taken immediately on waking and precisely timed — typically at waking, then 15, 30, and 45 minutes after — with no eating, brushing, or getting up to make coffee first. Even a 10-minute delay in that first sample can wipe out the very rise you're trying to measure. Expert consensus guidelines lay out strict protocols — exact timing, verified wake time (some studies use electronic monitors), no delays — because sloppy collection produces meaningless numbers (Stalder 2016). If you run this test, follow the instructions to the letter, and if you can't guarantee that precision, a standard four-point curve may be more reliable in practice.

5. 24-Hour Urinary Free Cortisol

Instead of a snapshot, this test pools all the free cortisol your kidneys excrete over a full day, giving an integrated measure of total daily production. You collect every drop of urine for 24 hours into a provided container.

It's a classic screen for Cushing's syndrome because it captures overproduction that a single blood draw might miss, and it's less affected by the minute-to-minute swings of a rhythm. Practically, you keep the collection jug refrigerated and capture every void across the full 24 hours; the lab often measures creatinine alongside to confirm the collection was complete. Its weakness is the mirror image of its strength: by averaging the whole day, it tells you the total but not the shape — you lose the timing information a salivary curve provides. Collection errors (missing a void, incomplete collection) are also common and can skew results, and very high fluid intake can dilute the sample, which is why results are usually interpreted against that creatinine check.

6. Dexamethasone Suppression Test

This is a dynamic test, not a static measurement. You take a small dose of dexamethasone (a synthetic glucocorticoid) at night, and cortisol is measured the next morning. In a normal system, dexamethasone signals the brain to switch cortisol production off, so the morning value should be suppressed to a low level. Failure to suppress suggests the axis isn't responding to normal feedback — a hallmark of Cushing's syndrome. In practical terms, it's like testing whether the brakes work: give the system a signal to stop, and see if it does. The most common version is the overnight 1 mg test, taken around 11 p.m. with the blood draw at 8 a.m. the next morning.

A few things can muddy the result and are worth flagging to your clinician: estrogen (from oral contraceptives or HRT) raises cortisol-binding globulin and can cause a falsely "non-suppressed" reading, and certain medications change how fast you clear dexamethasone. This is exactly why a single abnormal screen isn't a diagnosis. Comparative reviews of diagnostic accuracy show the overnight dexamethasone suppression test, late-night salivary cortisol, and 24-hour urinary free cortisol each perform well as first-line screens, which is why guidelines often recommend confirming an abnormal result with a second, different test (Galm 2020). No single test stands alone.

7. Hair Cortisol

The newest tool measures cortisol deposited in your hair as it grows, giving a retrospective window of chronic exposure — roughly one month of cumulative cortisol per centimeter of hair. Unlike saliva or blood, it's immune to the moment you happened to sample; it reflects your average stress load over months.

The practical appeal is obvious: it can't be thrown off by a single bad night, a stressful commute to the lab, or a mistimed sample. If you want to answer "has my body been under sustained stress load for months?", hair cortisol gets closer than any single-day test. Labs typically analyze the three centimeters of hair closest to the scalp, corresponding to roughly the last three months of growth.

Hair cortisol is still mostly a research and emerging-clinical tool, though, and factors like hair dyeing, heavy washing, bleaching, and even ethnicity can influence readings. A large meta-analysis has characterized both its stress-related signal and the many determinants that affect it, which is why it's best read as a chronic-load indicator that complements — rather than replaces — a rhythm test, not a stand-alone diagnosis (Stalder 2017).

How to Actually Test This (Most People Do It Wrong)

Here's the mistake almost everyone makes: they check cortisol once, with one method, at the wrong time, and then interpret a single value against a wide reference range. That approach can miss the exact rhythm distortions that explain how you feel. A perfectly normal 8 a.m. blood cortisol can coexist with a midnight value that should be near-zero but isn't — and that mismatch is invisible unless someone looks for it.

The root-cause approach does three things differently. First, it matches the test to the question — the single most important decision, because the right method depends entirely on whether you're screening for disease or investigating a rhythm: a late-night salivary or dexamethasone test to rule out true excess, but a four-point salivary curve (or CAR) when the complaint is fatigue, poor sleep, and stress, because those are rhythm problems. Second, it reads the pattern, not just whether each point falls inside "normal": a technically-normal morning value paired with a high midnight value is still a dysregulated curve. Third, it interprets your cortisol alongside your thyroid, blood sugar, and sex hormones, because these systems talk to each other — a flat cortisol curve in a perimenopausal woman with a sluggish thyroid is a very different story than the same curve in someone else.

There's also a practical layer most people miss: preparation. Things that transiently move cortisol — an intense workout the morning of the test, acute illness, alcohol the night before, shift work, and some medications — can distort a single snapshot. A good testing plan controls for those variables or interprets around them, which is another reason the person reading the result matters as much as the result itself.

This is the difference between "your labs are normal, everything's fine" and actually understanding why you feel wired, tired, and stuck. For the nutrition side of resetting a dysregulated curve, our guide to foods that lower cortisol covers what to eat to support a healthier rhythm.

Evidence-Based First Steps

If you want to check your cortisol levels the right way, here's a low-risk starting sequence:

• Match the test to the symptom. Suspect true excess (rapid central weight gain, purple striae)? A late-night salivary cortisol or dexamethasone suppression test is appropriate (Nieman 2008). Chronic fatigue and poor sleep? A four-point salivary curve maps the rhythm.
• Collect precisely. If you run a CAR, sample immediately on waking with no delays, per consensus protocols (Stalder 2016).
• Confirm abnormal results with a second, different test before acting on them (Galm 2020).
• Consider hair cortisol if you want a months-long average of chronic stress exposure (Stalder 2017).
• Never interpret a single number in isolation — read the curve, and read it alongside thyroid, blood sugar, and sex hormones.

The Bottom Line

Checking your cortisol levels isn't about getting one number — it's about capturing a rhythm and reading it in context. A morning blood draw has its place, but the tests that actually explain wired-and-tired symptoms are the ones that map the curve across the day (four-point saliva, CAR) or the ones that rule out true excess (late-night saliva, dexamethasone suppression, 24-hour urine). Hair cortisol adds a long-range view of chronic load. The right choice comes down to a single question: are you screening for a disease, or trying to understand a symptom pattern? Answer that first, and the correct test — and the correct time to run it — usually falls out naturally.

Just as important is resisting the urge to over-interpret a single result. Cortisol moves in response to sleep, illness, caffeine, exercise, and stress, so one out-of-range value on one day is a prompt to look closer, not a diagnosis. The reliable signal is a consistent pattern across the right test, confirmed and read in context.

If your results have been called "normal" while you still feel off, the problem may not be your cortisol level — it may be that no one has looked at the shape of your curve or connected it to the rest of your hormones. The most common gap isn't a missing test; it's a missing interpretation — someone to see that a normal morning value plus a high midnight value plus a sluggish thyroid tells a coherent story. A naturopathic or functional-medicine practitioner can order the right sample type at the right time and interpret the whole pattern together, turning a confusing single value into a clear map of what your stress system is actually doing — and, more importantly, a sequenced plan for resetting it.

This article is educational and not a substitute for individualized medical advice. Cortisol testing should be interpreted by a qualified clinician. Seek prompt in-person care for red-flag signs such as rapid central weight gain with a rounding face and wide purple stretch marks, severe unexplained muscle weakness, fainting, very low blood pressure with nausea and vomiting, or a crisis of profound weakness and dizziness — these can signal Cushing's syndrome or adrenal insufficiency and require urgent evaluation.