How to Test Cortisol Levels: The 7 Tests That Actually Reveal Your Rhythm

You feel wired at 10 p.m. and wrecked at 7 a.m. You crash mid-afternoon, wake at 3 a.m. with your heart pounding, and carry a layer of belly weight that no amount of dieting seems to touch. Somewhere along the way, someone told you it might be your cortisol. So you asked for a test, got a single blood draw at a random hour, and were told your levels were "normal." But you don't feel normal.

Here's the thing almost nobody explains: one cortisol number, pulled at one moment, tells you almost nothing. Cortisol isn't a static level like your cholesterol. It's a rhythm — a wave that's supposed to rise sharply in the first 30 minutes after you wake and taper down to near-nothing by bedtime. The question that actually matters isn't "is my cortisol high or low?" It's "is my curve the right shape at the right time?"

This guide walks you through how to test cortisol levels properly — the seven tests clinicians actually use, what each one captures, why a single reading fools so many women, and how to read your results like a root-cause practitioner would.

Why testing cortisol is different from almost every other lab

Most labs measure a substance that stays relatively stable across the day: your fasting glucose, your ferritin, your TSH. Cortisol breaks that rule completely. It follows a strong circadian rhythm driven by your hypothalamic-pituitary-adrenal (HPA) axis. Levels peak within about 30–45 minutes of waking — a spike called the cortisol awakening response (CAR) — then decline steadily throughout the day, bottoming out around midnight.

That means the timing of your sample is as important as the value itself. A cortisol of 15 µg/dL is reassuring at 8 a.m. and alarming at 11 p.m. Miss the timing and you get a number that's technically accurate and clinically meaningless.

For women, there's an added layer. Cortisol testing interacts with your menstrual cycle, oral contraceptives, pregnancy, and perimenopause — all of which shift the cortisol-binding globulin (CBG) that carries cortisol in your blood. Estrogen raises CBG, which raises total blood cortisol without changing the free, biologically active fraction. So a woman on the pill can look "high" on a blood test while her active cortisol is perfectly ordinary. This is exactly why free-cortisol methods — saliva and urine — often tell a cleaner story. If you want the lifestyle and nutrition side of the picture, our guide to foods that lower cortisol and how to build a cortisol-friendly diet pairs naturally with the testing roadmap below.

The goal of good cortisol testing is to reconstruct your curve, not to snapshot a single dot on it. Here are the seven ways to do that.

1. Morning serum (blood) cortisol — the standard, with a catch

A single morning blood draw between 7 and 9 a.m. is the most common cortisol test and the one your family doctor will reach for first. It's genuinely useful for one thing: screening for very high (Cushing-range) or very low (Addison-range) cortisol at its expected daily peak.

The catch is what it can't do. A morning serum cortisol measures total cortisol — roughly 90% of which is bound to CBG and albumin and biologically inert. It captures a single instant, so it tells you nothing about whether your rhythm flattens too early, whether your night-time levels stay elevated, or whether your morning surge is blunted. For the subtle, day-to-day HPA-axis patterns most stressed women are actually asking about, one morning number is the bluntest instrument in the kit.

Use it as a first-pass safety screen, not as your answer.

2. Late-night salivary cortisol — the most telling single sample

If you could run only one cortisol test, a late-night (around 11 p.m.–midnight) salivary sample would arguably be the most informative. Cortisol is supposed to be at its floor at that hour. When it isn't — when your night-time saliva cortisol is elevated — that's a strong signal your rhythm is disrupted or, in clinical extremes, that something is driving excess cortisol.

Saliva measures free cortisol, the fraction that actually enters your cells, which sidesteps the CBG problem that muddies blood tests in women on estrogen. Late-night salivary cortisol is one of the best-validated screening tools for cortisol excess: a large systematic review found it among the most accurate first-line tests for diagnosing Cushing syndrome, with sensitivity and specificity both in the strong range (Galm 2020). For the far more common "my rhythm feels off" scenario, a raised night-time value helps explain why you're tired-but-wired at bedtime.

3. The cortisol awakening response (CAR) — your morning surge

Within 30–45 minutes of opening your eyes, healthy cortisol jumps 50–100% above your waking value. That surge, the CAR, is your body mobilizing to meet the day. A flattened or absent CAR is associated with burnout-type exhaustion, chronic stress, and poor sleep; an exaggerated CAR can track with anticipatory anxiety.

Measuring the CAR properly is finicky, and most people do it wrong. It requires saliva samples at the moment of waking, then +15, +30, and +45 minutes — with precise timing, because even a few minutes of delay after waking wrecks the result. Expert consensus guidelines were written specifically because sloppy sampling made early CAR studies contradict each other (Stalder 2016), and those guidelines were formally re-evaluated and updated as the method matured (Stalder 2022). If a lab offers a "morning cortisol" without capturing waking-plus-timed samples, it is not measuring your CAR.

4. The four-point salivary diurnal curve — the rhythm map

This is the test that maps the whole wave. You collect saliva at four (sometimes five or six) points across a single day — typically waking, late morning, afternoon, and bedtime — and the lab plots your curve against the expected downward slope.

This is where the functional-medicine angle earns its keep. A four-point curve can reveal patterns a single draw can never show: a high morning that never comes down, a flat-line curve (little difference between morning and night, a marker of HPA dysregulation), a blunted morning with a paradoxical evening rise, or a curve that's simply shifted late. Because it uses free salivary cortisol, it's clean in women on the pill. And because it captures the shape, it maps onto how you actually feel across your day — which is the whole point.

5. 24-hour urinary free cortisol — the total daily output

A 24-hour urine collection measures the total amount of free cortisol your body excretes across a full day. Instead of a snapshot or a curve, it gives you an integrated volume — how much cortisol you produced over 24 hours.

Its strength is smoothing out the moment-to-moment noise; its weakness is that it discards timing entirely. A normal daily total can hide a badly-shaped curve, and a collection that misses even a few hours skews the result. It remains a workhorse for screening cortisol excess and is often run alongside late-night salivary cortisol so clinicians can cross-check total output against night-time floor. Think of it as the "how much," complementary to the curve's "when."

6. The overnight dexamethasone suppression test — the brakes check

This one tests your feedback loop rather than your baseline. You take 1 mg of dexamethasone (a synthetic steroid) at 11 p.m., then measure serum cortisol at 8 a.m. In a healthy HPA axis, the dexamethasone signals the brain to shut down cortisol production overnight, so your morning value should be suppressed to a low level. If it doesn't suppress, your feedback braking system isn't responding normally — a red flag for cortisol excess that warrants deeper endocrine work-up.

This is a diagnostic tool for suspected pathology (like Cushing syndrome), not a lifestyle-rhythm test. But it illustrates a principle worth internalizing: cortisol dysfunction isn't only about the number — it's about whether the regulation still works.

7. Hair cortisol — the long-game average

Every other test here captures hours or a day. Hair cortisol captures months. Cortisol is deposited into the hair shaft as it grows (roughly 1 cm per month), so a 3 cm segment of hair near the scalp reflects your average cortisol exposure over about three months — a retrospective, chronic-stress biomarker no blood or saliva test can provide.

Hair cortisol is powerful for the long view, but the research is nuanced: a recent large analysis found hair cortisol relates to psychosocial and biological factors in complex ways, and it's better read as a chronic-exposure average than a precise diagnostic (Boileau 2026). It's most useful when you want to know, "Has my body been under sustained load for the past season?" — not "what's my curve today?"

How to actually test cortisol (most people do it wrong)

Here's where root-cause testing diverges sharply from the standard 8 a.m. blood draw. If you want data that actually explains how you feel, the method matters more than the machine.

Test the rhythm, not a dot. The single most common mistake is accepting one blood cortisol as "your cortisol." It isn't. Prioritize methods that capture the shape of your day — a four-point salivary curve for lifestyle patterns, a late-night salivary sample as a high-yield single test. Reserve the single morning serum for safety screening at the extremes.

Match the method to the question. Free cortisol (saliva, urine) beats total blood cortisol for women, because it bypasses the estrogen-driven CBG shift that inflates total-blood readings on the pill or in pregnancy. If a clinician calls your blood cortisol "high" while you're on oral contraceptives, ask whether they've accounted for CBG — or better, retest with saliva.

Control the variables. Cortisol is exquisitely reactive. A recent professional-society consensus emphasized how much pre-analytic conditions — timing, recent stress, sleep, sampling technique — shape cortisol biomarker reliability, and pushed for standardized protocols precisely because sloppy collection produces noise (Fleseriu 2026). Before you sample: no eating, brushing teeth, or vigorous exercise in the 30 minutes prior to a saliva sample; avoid alcohol the night before; note where you are in your menstrual cycle; and if you can, sample on a typical day, not the morning of a flight.

Timing is the test. For the CAR, the clock starts the instant you wake — not when you get out of bed, not after coffee. A 10-minute delay can invalidate the result. For late-night saliva, hit the actual pre-sleep low, not "evening-ish." Precision in timing is what separates a meaningful cortisol curve from expensive confetti.

Interpret patterns together. A single flagged value means little. A flat curve plus poor sleep plus a blunted CAR plus an elevated hair cortisol tells a coherent story of chronic HPA load. Root-cause interpretation reads the constellation, not the star.

Evidence-based first steps

If you suspect your cortisol rhythm is off, here's a low-risk, sequential way to get real answers:

• Start with a late-night salivary cortisol or a four-point salivary curve, not a single random blood draw. These free-cortisol methods are the highest-yield first tests for mapping rhythm and screening excess (Galm 2020).
• If you use a CAR test, follow the consensus protocol exactly — sample at waking and at timed intervals, with strict adherence to timing, or the result is uninterpretable (Stalder 2016).
• Flag your hormonal context to whoever orders the test: menstrual phase, oral contraceptives, pregnancy, perimenopause. This determines whether blood or free-cortisol testing is appropriate.
• Standardize your collection day — typical sleep, no alcohol the night before, no exercise or food 30 minutes before saliva sampling — because pre-analytic conditions drive much of cortisol's variability (Fleseriu 2026).
• Consider hair cortisol only for the chronic-exposure question, understanding it's an average, not a same-day diagnostic (Boileau 2026).
• Address the obvious inputs first — sleep timing, caffeine after noon, blood-sugar swings, and evening light — all of which shape your curve before you spend a dime on testing.

The Bottom Line

Cortisol isn't a number — it's a rhythm, and the whole art of testing it well is capturing the shape of your day rather than a single dot on the curve. A random morning blood draw can miss a flat-lined curve, a blunted awakening surge, or an elevated night-time floor entirely. That's why so many women are told their cortisol is "fine" while they live the symptoms of a rhythm that clearly isn't.

The smarter path is method-matched to the question: free-cortisol saliva or urine over total blood, a four-point curve to map the shape, late-night saliva as a high-yield single test, and hair cortisol for the long-game average — all collected with the timing precision the science demands. And crucially, results are read as a pattern, not a single flag.

If your curve looks off, the next step is interpretation, not panic. A naturopathic or functional-medicine practitioner who can read these patterns together — alongside your sleep, cycle, and stress load — will get far more from your results than a lone lab value ever could. If you'd like help mapping which tests fit your situation, our care coordinator can point you toward a testing plan and a practitioner who reads the whole curve.

This article is educational and not a substitute for individualized medical advice. Cortisol testing should be interpreted by a qualified clinician. Seek urgent, in-person care for red-flag symptoms such as sudden severe weakness, dizziness or fainting, unexplained rapid weight loss with darkening skin, persistent vomiting, or severe abdominal pain — these can signal an adrenal crisis or other serious endocrine emergency.