How to take a hair sample for a cortisol test, step by step

To take a hair sample for a cortisol test, you cut a small lock of hair from the back of your head, just below the crown, as close to the scalp as you can. You mark which end was nearest the scalp, pack it as the kit shows and post it to the laboratory promptly. It takes a few minutes, needs no fasting and can be done at any time of day. The details matter more than they look, though: where you cut, how close to the skin you get and what you tell the laboratory all affect how well the result reflects the months you think you are testing.

What the sample needs to capture

Scalp hair grows 0.6 to 1.4 cm a month, about 1 cm on average, and takes up cortisol as it grows.¹ So the 3 cm nearest the scalp hold roughly the last three months, and 92 per cent of the studies examined in a systematic review used segments of 3 cm or less.² Every step below is about keeping that record intact: hair from the right place, cut at the right end and kept the right way round.

There is a built-in delay. New hair takes about two weeks to grow up to the surface, and scissors always leave a short stub behind. In one study, people following written and spoken instructions left about 0.8 cm of hair on the scalp, so the first centimetre of the sample reflected hair formed roughly 1.3 to 2.2 months before the cut.³ A hair sample is a record of the recent past, not of last week. If you want it to cover a particular period, the hair sample calculator works out the dates.

Step 1: Choose where to cut

Take the sample from the back of the head, just below the crown, an area researchers call the posterior vertex. Cortisol is not spread evenly across the scalp. In an early study, results varied by 30.5 per cent between samples from different parts of the head, but by 15.6 per cent between samples from the posterior vertex.⁴ That is why laboratories and research studies use the same area, and why repeat results are easiest to compare if you always cut from there.

A 2026 study compared one larger lock with three smaller locks taken close together and pooled. The pooled samples varied less within the same person for most hormones, and concentrations in single locks differed slightly with how high on the head they were cut.⁵ So if your kit suggests taking two or three thin locks from neighbouring spots along the same band of the head, that is the reason.

Step 2: Take enough hair

Follow your kit's instructions on how much hair to take. Typical kits ask for a small lock a few millimetres thick. Modern laboratory methods are sensitive, and one method measured seven steroid hormones from 10 mg of hair,⁶ but the laboratory also has to wash, dry and weigh the sample, so a lock that is too thin may not give a result.

Length matters as well. Most laboratories test the 3 cm nearest the scalp, so the hair needs to be at least that long. Longer hair is fine as long as the scalp end is clearly marked. Hair further from the scalp has been washed and exposed to light for longer and holds less cortisol, which is why the newest centimetres are the ones that count.⁷

Step 3: Cut as close to the scalp as you can

Lift the lock slightly and cut straight across with clean, sharp scissors, right against the skin. At about 1 cm of growth a month, every millimetre left on the head is roughly three days of the most recent record that the sample misses.¹ Some stub is unavoidable, but the closer the cut, the closer the sample gets to the present.

It is easier if someone else does this, because the back of the head is hard to see. Lift the top layer of hair out of the way and cut from underneath, so the hair around it covers the spot.

Step 4: Mark the scalp end, pack it and post it

The laboratory needs to know which end grew most recently. In studies that cut hair into 3 cm sections, cortisol fell by 30 to 40 per cent from one section to the next moving away from the scalp.⁷ A lock analysed the wrong way round would therefore read low and describe the wrong months.

Keep the lock lying straight, with the cut ends together. Mark or secure the scalp end exactly as your kit shows, usually with a clip, tape or a printed arrow on the foil or envelope supplied. Make sure the hair is completely dry before you seal it.

Hair does not need a fridge or freezer, which is one reason it suits home testing. It is not completely stable, though. In a study that analysed the same people's hair repeatedly during storage, cortisol fell steadily over two years and agreement with freshly analysed samples weakened, and the authors recommend analysis ideally within three to six months of sampling.⁸ Post your sample within a few days rather than leaving it in a drawer, and fill in the request form at the same time.

What to avoid beforehand, and what to tell the laboratory

You do not need to fast, avoid exercise or cut at a particular time. Hair cortisol was unrelated to the time of day of sampling in one study, and sweating heavily in a sauna or on a treadmill did not change it.⁹ A few other things can shift the result, though:

• Dye and heat treatments. In the Whitehall II study, people who dyed their hair had 12.5 per cent lower hair cortisol,¹⁰ and heat-based treatments were linked to lower values in a South London study.¹¹ In the early method study, colouring before sampling lowered cortisol but dyeing the hair after it had been cut did not.⁴ Studies show an association rather than a safe waiting period, so avoid new treatments in the weeks before you cut if you can, and tell the laboratory about any in the past few months. See testing dyed or treated hair.
• Steroid medicines. In the Lifelines cohort, people using only inhaled, nasal or skin corticosteroids had lower hair cortisone (6.8 against 8.5 pg/mg),¹² while systemic corticosteroid use was linked to higher hair cortisol in Whitehall II.¹⁰ Do not stop a prescribed medicine for a test. Tell the laboratory what you use, including steroid creams or lotions on the scalp.
• Washing. How often people washed their hair made no difference to the sections nearest the scalp in one study, only to the third 3 cm section.¹³ Wash as you normally do.

Common mistakes

• Cutting from the top or sides of the head instead of the back, just below the crown.
• Cutting partway down the hair rather than at the scalp, which drops the most recent weeks.
• Losing track of which end was nearest the scalp.
• Taking a lock that is too thin or too short to analyse.
• Using hair from a hairbrush or a haircut. Neither was cut at the scalp from one spot, so neither represents the most recent months.
• Sealing hair that is still damp.
• Leaving the sample for weeks before posting it.
• Forgetting to mention dye, heat treatments or steroid medicines.

Very short hair, grey hair and body hair

If your hair is shorter than 3 cm, the sample covers less time, because each centimetre is roughly a month.¹ Ask the laboratory what minimum length it accepts before ordering. If your hair is thinning, pooling two or three small locks from neighbouring spots can help; the guide to cortisol and hair loss covers this.

Grey hair is fine. A study comparing pigmented and grey hairs from the same people found no difference between them.¹⁴ Body hair is a different matter: in the same study most hormones correlated only weakly between scalp and body hair, and the authors concluded body hair is not a suitable substitute when scalp hair is not available.¹⁴

Once the sample is posted, the guide to hair cortisol results explains how to read the report, and how often to test covers when a repeat sample is worth taking.