Every human study of rooibos we could find
Eight, spanning forty-six years. Read the sample sizes: they are the part that usually goes missing when these get retold.
Iron absorption — 1979
Volunteers swallowed a measured dose of iron and then drank rooibos, ordinary tea or boiled water. Absorption was measured two weeks later on a whole-body counter. Rooibos 7.25%, ordinary tea 1.70%, water 9.34%.
Ordinary tea reduced iron absorption strongly and significantly. Rooibos did not reduce it significantly. This is the origin of every “rooibos does not block iron like tea does” line you will read, and it is a real result.
Small, old, all-male, and never replicated that we can find. Still the only direct test of the question.
ACE and nitric oxide — 2010
One 400 ml serving of rooibos, green tea or black tea. Rooibos significantly inhibited angiotensin-converting enzyme activity at thirty and sixty minutes. Nitric oxide did not change.
Seventeen people, one cup, one afternoon. It shows rooibos does something measurable in blood chemistry. It does not show that anything happens to a person.
Rehydration after exercise — 2010
Athletes dehydrated by 3% of body mass, then rehydrated with rooibos, bottled water or a carbohydrate drink.
Rooibos was no more effective at rehydrating them than plain water. The carbohydrate and electrolyte drink beat both.
A clean negative result on a claim rooibos is still sold with. Worth more than most of the positives, because somebody went looking and did not find it.
Blood lipids and oxidative stress — 2011
The strongest human result rooibos has. After six weeks, LDL cholesterol and triglycerides were significantly lower, HDL higher, markers of lipid peroxidation down, and the reduced-to-oxidised glutathione ratio up.
The same study's other finding is the one that never gets quoted: plasma antioxidant capacity was not altered. Total polyphenols in the blood went up; the blood's measured antioxidant capacity did not follow.
Forty people, no placebo group, a control period instead. Genuinely the best evidence there is, and it is a long way from proof. Everyone in it was already at cardiovascular risk, so it says nothing about a healthy person.
Cortisol and cortisone — 2014
In the human arm of a paper that also used rats and cells, rooibos consumption raised plasma cortisone in men and lowered the cortisol-to-cortisone ratio in both men and women.
This is the source of “rooibos lowers your stress hormone”. What it measured is a ratio between two related steroids in blood, not stress, and not how anyone felt. No sleep or anxiety outcome was measured.
Kidney stones — 2016
Rooibos and Japanese green tea were each given for a month, with urine analysed for the physicochemical and oxidative factors that drive stone formation.
Nothing significant changed in any of the risk factors the researchers measured. They describe their own study as a pilot needing much larger samples.
Eight people. Both the rooibos claim and any reassurance about it rest on almost nothing.
Heart structure, by ultrasound — 2024
The first properly controlled rooibos trial of any size. Adults at cardiovascular risk took capsules of standardised green or fermented rooibos extract — not tea — for twelve weeks, with the heart imaged by echocardiography before and after.
Some measurements moved. Left ventricular mass fell in the fermented group. But the interventricular septum also decreased significantly in the placebo group, which is exactly why placebo groups exist.
Much better designed than what came before, and about an extract in a capsule. If you are reading it as a reason to drink tea, you are reading past what was tested.
Exercise and oxidative stress — 2025
The most recent. Its actual purpose was to validate a way of inducing oxidative stress through exercise; rooibos polyphenols were the intervention used to test the model.
Markers of oxidative damage rose after exercise irrespective of what the men had taken. The authors' conclusion about rooibos is carefully worded: it did not complicate the onset of oxidative stress.
Not a finding that rooibos protects against exercise stress. The opposite of one, if anything.
Three of the eight found nothing, and that matters more than it looks
Rehydration: no better than water. Kidney stones: nothing moved. Exercise oxidative stress: markers rose regardless. And inside the best study, the antioxidant capacity measurement — the single thing rooibos is most often sold on — did not change.
Those results are almost impossible to find on the web, because there is nothing to sell in them. But three null results out of eight attempts is a real feature of this literature, and a reader deciding what to believe needs it more than they need another list of nine benefits.
It also tells you something useful about the good result. The 2011 lipid finding is more credible because the same field has produced clean negatives. A research area where everything works is a research area that is not really testing anything.
- Found something — 5
- Found nothing — 3
- Had a placebo group — 2
- Tested tea rather than extract — 6
- Known to include women — 1
- Known to be men only — 2
- Longer than six weeks — 1
The claims with no human evidence behind them
These are the things rooibos is most commonly sold for. In each case we searched and found no human trial. That is not the same as saying the claim is false — it means nobody has tested it in people.
| The claim | Human trial? | What actually exists |
|---|---|---|
| Helps you sleep | None found | Rooibos is caffeine-free, which is a fact about the leaf. No trial has measured sleep |
| Lowers blood sugar / helps diabetes | None found | A 2018 meta-analysis pooling 12 studies in 88 diabetic rodents |
| Helps weight loss | None found | Cell and animal work on fat metabolism |
| Reduces anxiety or stress | None found | The 2014 study measured a cortisol ratio in blood. Nobody was asked how they felt |
| Good for skin, taken as a drink | None found | One 28-day trial of a cream containing rooibos mixed with green tea |
| Eases colic in babies | None found | Nothing. See breastfeeding and babies |
| Relieves allergies or eczema | Unread | A 1982 paper, “The effect of rooibos tea on the type I allergic reaction”, exists. PubMed carries no abstract for it and we have not read the full text, so we cannot tell you what it found |
| Lowers blood pressure | None found | The single-dose ACE enzyme study in 17 people. Blood pressure itself was not the outcome |
“No human trial” is a statement about the literature, not about your body. Plenty of ordinary true things have never been through a trial. What the table rules out is the confident claim, not the possibility.
Why nearly all of this is rats and cells
Because trials are expensive and rooibos is a commodity crop. Nobody holds a patent on a leaf, so there is no company that can recover the cost of a large trial by selling the result. The research that does happen is mostly publicly funded, mostly South African, and mostly done at the scale a university laboratory can afford — which is cells, then rodents, and occasionally forty volunteers.
That is not a criticism of the scientists. The work at Stellenbosch, the Cape Peninsula University of Technology and the Medical Research Council is careful and it is honest about its own limits; the papers say “further studies are required” because further studies are required. The dishonesty enters later, when a rodent result is retold on a website as a benefit for you.
It also explains the shape of the literature. Rooibos is well studied at the level of what molecules are in it and what they do to a cell, and barely studied at the level of what happens to a person who drinks it. Both of those things can be true at once, and most writing about rooibos quietly swaps the first for the second.
What can be said about rooibos without any argument
Composition is not effect, and composition is checkable. These are the statements that need no hedging:
- It is caffeine-free. Not low in caffeine — free of it, per the South African National Biodiversity Institute. If you are avoiding caffeine, that is a real reason to drink it and it is not a health claim.
- It is not tea. Rooibos is Aspalathus linearis, a legume from the Cape fynbos. Tea is Camellia sinensis. The full comparison.
- It contains a distinctive polyphenol, aspalathin, found in almost nothing else, which is why the laboratory literature on it is as large as it is.
- It contains small amounts of iron, magnesium, phosphorus, sodium, chloride and potassium, and ascorbic acid at levels that vary with how you brew it. Measured in 1979, in the same study as the iron result.
- It grows commercially in one region on earth. That one is ours to talk about. The plant and the place.