Rosemary oil for hair growth, examined honestly

Rosemary oil is having a very long moment. It is presented on social media as a natural equivalent to minoxidil, usually with before-and-after photographs and rarely with a citation.

There is, unusually for this genre, a real study behind it. It is worth reading properly, because what it shows is more interesting than either the hype or the dismissal.

The rosemary oil for hair growth trial

In 2015, a randomised comparative trial in Iran assigned around 100 people with androgenetic alopecia to either rosemary oil or 2% minoxidil, applied to the scalp for six months. Hair counts were measured photographically at baseline, three months and six months.

The findings, in brief: neither group changed much at three months. Both groups showed a significant increase in hair count at six months, and the difference between them was not statistically significant. Scalp itching was reported more often in the minoxidil group.

That is a genuine result from a genuine randomised trial, and anyone who tells you there is no evidence for rosemary oil has not read it.

What the trial does not show

Now the qualifications, all of which matter.

There was no placebo arm. The study compared two active treatments. Without a group receiving carrier oil alone, you cannot separate the effect of rosemary from the effect of applying something to the scalp and massaging it in daily for six months — and massage itself has at least a plausible mechanism. It is entirely possible that a portion of the improvement in both arms came from the ritual rather than the molecule.

It is one trial. Single studies in any field are provisional by default. The finding needs independent replication in other populations before it becomes knowledge rather than a lead.

It was in androgenetic alopecia specifically. Whether rosemary oil does anything for telogen effluvium, postpartum shedding, or generally sluggish growth was not tested. Most people using it are not in the population that was studied.

Photographic hair counting has limits. It is the standard method and it is reasonable, but it is sensitive to lighting, positioning and observer effects in ways that make small differences hard to trust.

2% minoxidil is the lower-strength formulation. Comparable to the weaker of the two standard concentrations, not to 5%.

The honest position

Rosemary oil is a low-risk thing to try, with one real supporting trial, a plausible mechanism, and no way to know yet whether it does anything meaningful for you specifically.

That is genuinely a better position than almost anything else in the natural hair-growth category, most of which has nothing at all. It is also a long way from what the videos claim.

If you want to try it: dilute it properly — essential oils are not skin products at full strength — apply to the scalp rather than the lengths, use it consistently, and evaluate at six months with photographs taken in the same light, because the growth cycle guarantees nothing will be visible sooner.

If your shedding is sudden, patchy, or accompanied by scalp symptoms, none of this applies and you should see a doctor. And if you have androgenetic hair loss and are choosing between a treatment with one trial and a treatment with thirty years of data, that is a conversation with a clinician rather than a comment section.

The wider lesson

The reason to write this at length is not rosemary. It is the shape of the argument, which repeats constantly in hair care.

A single study appears. Its result is real but narrow. The narrowness is stripped off in the retelling. Within a year it is "clinically proven", it appears on packaging, and anyone pointing out the missing placebo arm is treated as a cynic.

The test to apply, to this and everything else: how many trials, how many people, compared with what, measured how, and over how long? Ask it of rosemary oil and you get "one, a hundred, minoxidil, photographs, six months" — which is enough to be interesting and not enough to be sure. Ask it of most hair supplements and you get silence.