Cortisol Testing: Which Ones Are Worth Paying For

Somewhere between the fatigue and the 4pm collapse, you decided you wanted your cortisol measured. Then you discovered there are four ways to do it, three of them can be ordered online, and one costs £180 and arrives with a chart shaped like a hill, and an invoice shaped like a small emergency.

Cortisol testing is genuinely useful in a narrow set of circumstances and close to meaningless in the rest. Knowing which situation you are in saves a lot of money.

The four methods, honestly graded

Morning blood cortisol. Evidence: Strong, for what it is designed to do. This is the standard clinical test, taken around 8–9am when levels peak. It is the right test for suspected Addison’s disease or Cushing’s syndrome, and it is usually a starting point rather than a final answer.

Late-night salivary cortisol. Evidence: Strong, in the right context. Cortisol should be at its lowest around midnight. A raised late-night result is a well-validated screen for Cushing’s, and it is ordered by endocrinologists for that reason.

Four-point salivary curve. Evidence: Mixed. This is the one sold direct to consumers. Measuring the daily rhythm is a legitimate research method, but the reference ranges are poorly standardised between labs, results vary substantially day to day, and there is no established treatment protocol attached to a wobbly curve.

Hair cortisol. Evidence: Emerging. Interesting for research into long-run stress exposure. Not yet useful for making a decision about your Tuesday.

The problem is not the sample, it is the interpretation

Cortisol is meant to move. It peaks within about 30 minutes of waking, falls across the day, and bottoms out around midnight. It also rises with a poor night’s sleep, an argument, a hard workout, or the mild dread of doing a saliva test correctly.

So a single reading is a snapshot of one moment, not a verdict on your endocrine system. Results sold as evidence of “adrenal fatigue” are the clearest example — a systematic review of dozens of studies found no consistent evidence that the condition exists as described — which has not slowed down the people selling the cure for it in the slightest. Your symptoms are real. That particular explanation for them is not.

A wobbly curve on an expensive chart tells you that cortisol varies. Cortisol is supposed to vary.

When cortisol testing is genuinely warranted

Ask your doctor about proper testing if you have unexplained low blood pressure, dizziness on standing, salt cravings, darkening skin or persistent vomiting — the picture that raises the question of adrenal insufficiency.

Or if you have rapid central weight change, purple stretch marks, easy bruising, muscle weakness and high blood pressure together — the picture that raises Cushing’s.

Both are uncommon. Both are serious and treatable. Neither is diagnosed from a home saliva kit, and both deserve a clinician rather than a chart.

What I would spend the money on instead

Nothing, first. The interventions for a stressed system are the same whether your cortisol is measured or not, which is the quiet reason the test rarely changes anything.

Sleep, protected properly. Sleep debt raises cortisol more reliably than most things you can address.

Training load, adjusted. If you are grinding through high-intensity sessions on poor sleep, you are adding to the pile.

Eating enough. Under-eating is a stressor, and it is the most commonly missed one.

Five minutes of extended-exhale breathing. Free, evidenced, faintly ridiculous, works.

And if you are considering supplements for this, adaptogens are worth reading about first — some have real data, most have marketing.

Questions people actually ask

Are at-home cortisol tests accurate?

The measurement itself is usually reliable, but reference ranges vary between labs and single readings fluctuate a great deal. The weakness is in the interpretation, not the sample.

Is adrenal fatigue a real diagnosis?

No. A systematic review of the published studies found no consistent evidence supporting it. The symptoms people describe are real and worth investigating, but this explanation is not recognised clinically.

What time should cortisol be tested?

Standard blood testing is done between 8 and 9am, when levels peak. Late-night salivary testing is done around midnight, when they should be at their lowest.

More in the stress section.