Updated 2 September 2026. Test carried out May 2024; photographs unchanged. What we have added is the chemistry behind the colour change, and a clearer account of what this test did and did not measure.
Hibiscus is in 13 of our 24 shades, and it is the only herb we use whose colour changes depending on what you mix it with. Add water and you get one colour. Add lemon juice and you get another — from the same powder, in the same hour, on the same hair.
We ran it on white hair swatches so the difference would be visible. Here is what happened, and why.
The test
Organic hibiscus flower powder, mixed two ways: once with plain water, once with lemon juice added. One hour on white hair swatches, then rinsed.
Mixed with water: blue-purple
Mixed with lemon juice: red-purple
Side by side
Water: blue-purple. Lemon: red-purple. Same powder, same hour, same hair.
Why the colour changes
You have probably seen this happen before, in a kitchen. Squeeze lemon onto red cabbage and it turns bright pink. Add something alkaline and it goes blue-green. The cabbage has not been damaged and nothing has been added to it — the pigment simply looks different depending on how much acid is around it.
Hibiscus carries the same family of pigments: anthocyanins. Blackcurrants, red wine, blueberries and red cabbage all owe their colour to them.
What makes anthocyanins unusual is that the molecule changes shape with acidity, and each shape is a different colour. In acid conditions one form dominates and reads red. As acidity falls, the molecule loses a proton and rearranges into another form, which reads purple. Less acid still and it shifts again, towards blue.
Nothing is created or destroyed at any point. It is one pigment, sitting in different arrangements, and acid decides which arrangement you are looking at. That is the whole of what the lemon juice did.
So why was the water version blue rather than neutral purple?
This is the part we found most interesting, and it is not our own finding.
A 2018 study at the University of Leeds looked at anthocyanins from blackcurrant applied to hair — the same pigment family, a different fruit. They found that the pigment does not bond to hair the way henna does. It adsorbs onto the surface, and the pigment molecules stack on one another as more arrives.
They also found something that bears directly on our swatches: hair appears to stabilise the blue form of the pigment. The suggestion is that positively charged ions present in hair, together with the pigments stacking against each other, hold the molecule in the arrangement that reads blue.
If that holds for hibiscus too, our water swatch was not neutral — the hair itself was pushing the colour towards blue. The lemon pushed back the other way. That would explain why the difference between the two swatches was as clear as it was.
What we did not measure
We would rather list this than let the photographs imply more than they show.
- We did not measure the pH of anything. Not the lemon juice, not the two pastes, not the hair afterwards. We know one was acidic and one was not, and that is all we can say with any precision. Lemons vary, and so does the ratio you happen to mix at.
- We did not test how long either colour lasts. One hour on, rinsed, photographed. No wash-fastness testing at all. Anthocyanins are not bonded to the hair, so we would expect them to fade sooner than henna — but expecting is not measuring.
- White hair only. White swatches were chosen precisely because nothing hides the result. On pigmented hair there is far less to see, and on dark hair likely nothing at all.
- One batch, one hour, one pair of swatches. Not a controlled trial.
- We have not measured how much pigment stays on the fibre. Nobody has, for hibiscus specifically. The blackcurrant study is the closest thing that exists, and its authors hold commercial interests in that work.
One thing we used to say, and no longer do
An earlier version of this article said that the particle size of the powder was the key to the result, and that milling it too fine would destroy the pigment. We cannot support that. A powder does need to be fine enough to disperse smoothly — a lump colours nothing but itself — but we have no evidence that fine milling damages anthocyanins, and we should not have written it as though we did.
The same version described the pigment as colouring the hair from within. That is not what the evidence shows either. Anthocyanins sit on the surface of the strand. We have corrected both.
What this means if you use a Radico shade
Do not expect purple. Hibiscus is in 13 of our 24 shades, but in almost all of them it sits near the bottom of the ingredient list — a small share, working underneath henna and indigo rather than in place of them. The one exception is Honey Blonde, where it comes first.
What the pH behaviour does explain is something we say elsewhere on this site. Amla is in 23 of our 24 shades, and amla is acidic. It keeps the paste on the acid side — which, as these swatches show, is the side hibiscus reads red on. Take the acid away and the same pigment would drift the other way.
If you want the purple itself, that is what the single herb is for. Mixed with water it leans blue; with lemon juice or apple cider vinegar, red. Strand test first — this is a pigment that behaves differently depending on what is already on your hair and what you mix it with.
Reference
Rose PM, Cantrill V, Benohoud M, Tidder A, Rayner CM, Blackburn RS. Application of anthocyanins from blackcurrant (Ribes nigrum L.) fruit waste as renewable hair dyes. J Agric Food Chem. 2018;66(26):6790–6798. doi:10.1021/acs.jafc.8b01044 · PMID 29808681.
Blackcurrant, not hibiscus — the same pigment class with a different sugar attached. Two of the authors declare interests in a company commercialising the technology, which is worth knowing when you read it.
Organic Hibiscus Flower Powder →
Radico Organic Hibiscus Flower Powder is a cosmetic product for external use on hair. Patch test on the inner forearm 48 hours before a full application.