Claim as made
Clinically proven to match minoxidil for hair regrowth — the natural alternative, without the side effects.
The market's own words. No brand is named, because the claim is the thing under examination and it belongs to dozens of them.
Rosemary oil works as well as minoxidil
One 2015 trial compared rosemary oil with minoxidil 2%, not the 5% most people buy, and had no placebo arm. Nothing has replicated it.
Claim as tested
In adults with androgenetic alopecia, does topical rosemary oil, compared with placebo, increase hair count at six months — and separately, does it match minoxidil 5%, the concentration most people actually buy?
The same claim written as a proposition somebody could go and check: who, compared with what, measured how, over how long.
What kind of missing this is
Claim inflation
The study found something. The claim describes something larger, or faster, or more certain. The distance between the two is usually the marketing budget.
What the strongest study actually found
Panahi and colleagues, in Skinmed in 2015, randomised people with androgenetic alopecia to rosemary oil or to minoxidil, and reported comparable hair counts at six months. That is the entire human evidence base. One trial, eleven years old, in a low-impact journal, and it is the source of every "clinically proven" on every bottle.
Two details in it matter more than the headline, and both are in the paper rather than hidden.
The comparator was minoxidil 2%. The formulation most people buy, and the one most likely to be sitting in the bathroom of anyone reading this, is 5%. A trial showing you match the weaker preparation is not a trial showing you match the one on the shelf.
And there was no placebo arm. No vehicle control, nothing inert. So the trial can support a statement about rosemary oil relative to minoxidil 2%, and it cannot support a statement about rosemary oil relative to nothing at all — which is the comparison a person deciding whether to buy it is actually making. Both arms improved. In a six-month scalp trial with no control, so does massaging your head with olive oil, and nobody in this study was assigned to find out.
Eleven years is a long time for a result this commercially useful to go unreplicated in humans.
What would change this verdict
A randomised, vehicle-controlled trial in at least 120 adults with androgenetic alopecia, running six months, with three arms: rosemary oil, minoxidil 5%, and the vehicle alone. Primary endpoint hair count per square centimetre by standardised macrophotography, assessed by someone blinded to the arm.
The vehicle arm is the part that does the work. Without it, the 2015 result cannot distinguish "rosemary oil grows hair" from "six months of scalp massage with any oil, in people who have started paying attention to their hair, produces a measurable count difference."
A positive result against vehicle moves this to Partially established immediately. A positive result against minoxidil 5% moves it to Established, and I would say so in the same tone as everything else here, because it would be a genuinely good outcome — a cheap, tolerable option for a condition where the licensed one irritates a lot of scalps.
What this does not mean
It does not mean rosemary oil does nothing. The 2015 trial is real, it was randomised, and both arms improved. It means the trial cannot tell us how much of that improvement belongs to the rosemary.
It does not mean minoxidil is better. Nobody has run the head-to-head against 5% that would answer that.
It does not mean people reporting regrowth are mistaken. Androgenetic alopecia fluctuates, shedding is seasonal, and anybody who starts a new hair routine also starts photographing their scalp in better light. Those are not lies; they are the reasons control arms exist.
And it does not mean essential oils are unsafe. Rosemary oil is a known contact sensitiser in some people, which is a specific and manageable risk, not a general warning.
If you are deciding whether to buy
You are paying for an essential oil, typically at a few pounds or dollars a bottle, which is roughly a tenth of what a year of minoxidil costs. At that price the calculation is not really about evidence at all — it is about whether you would rather spend six months on the cheap option that has one unreplicated trial, or the expensive one that has many.
What you are not paying for is a proven equivalent to the treatment you have probably already been offered. The trial that phrase rests on used half the strength.
The thing actually worth knowing is that androgenetic alopecia is progressive, and six months spent on an option that turns out not to work is six months of loss you do not get back. That is the real cost here and it is not on the label. If your hair loss is early and slow, the low-stakes experiment is defensible. If it is moving fast, the cheap bottle is an expensive way to buy time you do not have.
Sources
Type and funding are stated because they change what a study is allowed to prove, and because a trial paid for by the company selling the thing is not disqualified — it is just not the same evidence as one that was not.
- Panahi Y, Taghizadeh M, Marzony ET, Sahebkar A. Rosemary oil vs minoxidil 2% for the treatment of androgenetic alopecia: a randomized comparative trial. Skinmed. 2015 Jan-Feb;13(1):15-21. PMID: 25842469. Source
- PubMed and ClinicalTrials.gov searches, 3 September 2026, for human trials of Rosmarinus officinalis in alopecia. No replication of the 2015 trial identified. Source