Rosemary Oil vs Minoxidil: What the 2015 Trial Actually Showed

Key Takeaways

If you have read that rosemary oil "works as well as minoxidil", you have read a summary of one paper published in Skinmed in 2015. It is a real trial, it was randomised, it ran six months, and its result is honestly stated in its own abstract. Almost everything else attached to it since was added by people who did not read the methods.

This article walks through what the trial did, what it found, what it could not find because of how it was built, what the mechanism literature does and does not add, and how a person who wants to use rosemary oil anyway should do so without harming themselves or delaying something that works.

The Design

The trial enrolled 100 patients with a clinical diagnosis of androgenetic alopecia and randomised them to two arms of 50: rosemary oil applied to the scalp, or minoxidil 2% solution. Treatment ran six months. After a baseline visit, patients returned every three months for efficacy and safety assessment.

The primary efficacy measure was hair count by standardised professional microphotography of a target area, taken at baseline, three months and six months. Safety and tolerability were captured as the frequency of dry hair, greasy hair, dandruff and scalp itching at each visit.

Three details of the design carry all the weight in what follows. The comparator was minoxidil at 2%, not 5%. There was no placebo or vehicle arm. And the published abstract does not describe blinding of patients or of the photographic assessors.

What It Found

Three findings, then, in the authors' own framing. Nothing happened at three months in either arm. Both arms increased hair count at six months. The arms did not differ. The trial "provided evidence with respect to the efficacy of rosemary oil", and rosemary caused less itching.

Every one of those statements is correct. The problem is what each one can bear.

The Three-Month Null

The fact that neither arm moved at three months is often skipped, and it should not be. Minoxidil 2% is an established active treatment; if it produced no measurable hair-count change at three months in this cohort, the trial's measurement was either too coarse or too early to detect a real effect — or the cohort's response to minoxidil was small.

Either way, the three-month null tells you the study was not sensitive to modest changes. A trial that cannot detect its own active comparator working at three months is a trial in which "no difference between arms" at six months is a weak statement.

The Missing Placebo Arm

This is the limitation that decides everything.

An active-comparator design answers the question are these two treatments similar? It answers does either of them work? only if the comparator's effect size is already firmly known at the dose and duration used, so that "similar to it" can be translated into "as effective as a known quantity".

Minoxidil 2%'s effect over six months in a cohort like this one is not a known quantity. It is real — the pivotal three-arm trial showed 2% beating placebo over 48 weeks — but its magnitude at six months varies by cohort and by measurement. Without a vehicle arm in this trial, two explanations fit the data equally well:

1. Rosemary oil is as effective as minoxidil 2% at six months.
2. Neither arm did much; the six-month rise in both groups reflects regression to the mean, seasonal hair-cycle variation, measurement drift, or the natural fluctuation of counts in patients who enrol when their loss feels worst.

The trial cannot distinguish these. Nobody reading it can. And the second explanation is not a straw man: patients seek treatment at a low point, hair counts vary with season and with the shedding cycle, and both arms increasing by a similar amount from a similar baseline is exactly what a null result looks like when there is no control group to anchor it.

The 2% Problem

Suppose you accept explanation one. Rosemary is then as good as minoxidil 2%. How good is that?

The pivotal 48-week trial of 393 men compared minoxidil 5%, minoxidil 2% and placebo. Five percent was clearly superior to 2% on non-vellus hair count, patient-rated coverage and investigator-rated coverage, producing 45% more regrowth than 2% at week 48, with an earlier onset of response. Five percent is the standard concentration in men for that reason, and it has been for two decades.

So the ceiling of the rosemary claim, granting every benefit of the doubt, is: equivalent to the weaker concentration of the second-line drug. That is not nothing — 2% is still better than placebo — but it is a long way from the phrase "rosemary oil is as good as minoxidil", which people hear as "as good as the product in the pharmacy". The product in the pharmacy is 5%.

And above minoxidil sits finasteride. In 1,553 men over two years, finasteride 1 mg increased vertex hair count by 138 hairs relative to placebo, while placebo progressively lost hair. Rosemary has never been within sight of that comparison.

The Other Limitations

Never replicated. More than a decade later, there is no second human trial of rosemary oil in androgenetic alopecia against any comparator. A 2025 network meta-analysis of male AGA treatments ranked topical rosemary among the over-the-counter products it compared with conventional drugs — which is a sign that researchers are trying to place it, not that the placing has been done, and that analysis still rests on the same single trial for its rosemary node.

Blinding. The abstract does not describe blinding. An oil and an alcoholic minoxidil solution look, feel and smell different; a patient knows which arm they are in, and unless the photographic assessors were blinded, so do they. Unblinded outcome assessment in a trial whose primary outcome is a count from a photograph is a real vulnerability.

Effect size. The abstract reports significance, not magnitude. It does not tell you how many hairs per square centimetre either arm gained. "Significant increase" at P < 0.05 in a sample of 50 can be a small number, and in the absence of a placebo arm, a small number is exactly what regression to the mean would produce.

Population. Patients with a clinical diagnosis of androgenetic alopecia — no severity grading is reported in the abstract, and no subgroup analysis by stage. Whether rosemary did anything for a Norwood V vertex or only for early thinning is unknown.

What the Mechanism Literature Adds

The case for rosemary is not built on the human trial alone; it is propped up by a body of laboratory work that is real and frequently misdescribed.

The key study applied rosemary leaf extract at 2 mg per day to the shaved backs of mice. In C57BL/6 mice whose hair regrowth had been interrupted by testosterone injection, the extract restored regrowth; in shaved C3H/He mice, it promoted growth. In a cell-free assay the extract inhibited testosterone 5-alpha-reductase by 82.4% at 200 µg/mL and 94.6% at 500 µg/mL, and activity-guided fractionation identified 12-methoxycarnosic acid as the responsible constituent. The extract and the isolated acid also inhibited androgen-dependent proliferation of prostate cancer cells by roughly two-thirds at 5 µg/mL and 5 µM.

Three things need saying about that.

First, it is leaf extract, not essential oil. The diterpenes doing the work — carnosic acid and its methoxy derivative — are non-volatile and are present in the steam-distilled oil only in small, variable amounts. The human trial used the oil. The mechanism was demonstrated in the extract. Nobody has measured how much of the active fraction a diluted oil delivers.

Second, the concentrations are test-tube concentrations. Ninety-four percent inhibition at 500 µg/mL describes an enzyme in a dish. Whether a 3% oil dilution achieves any meaningful 5-alpha-reductase inhibition in a human dermal papilla is unmeasured.

Third, a 2025 mechanistic review of carnosic acid in alopecia — which catalogues its antioxidant, anti-inflammatory and anti-androgenic activity — also states plainly that its poor solubility and instability in conventional formulations limit clinical application. That is the compound's own advocates noting that getting it into a scalp is an unsolved problem.

How One Trial Became a Movement

A 2023 review of topical hair-loss remedies on TikTok — rosemary oil, rice water, onion juice, garlic gel — set out to trace the evidence behind claims patients bring to clinic, and rosemary is the one with an actual randomised trial behind it. That is precisely why it spread. "Matches minoxidil in a randomised trial" is a true sentence, it is short, and it does not require the reader to know what 2% means or what a placebo arm is for.

The chain is easy to reconstruct. The abstract says no difference from minoxidil 2%. A summary drops the "2%". A video drops "no difference" and says "as effective as". A comment section adds "without the side-effects". Within a few iterations the trial has been recruited to support a claim — rosemary replaces minoxidil — that it was never designed to test and could not have tested if it had been.

None of that is the trial's fault. The authors compared two active arms, reported that they did not differ, and noted that rosemary caused less itching. The overreach is downstream.

The Comparison That Should Be Made

Read across the top row and the claim collapses to its real size: a single trial with no anchor, sitting under a two-decade pyramid of placebo-controlled data for the two drugs. Read down the side-effect column and rosemary's genuine advantage appears — it is gentler on the scalp — along with its own risk, which the internet omits: allergic contact dermatitis to rosemary, including to carnosol specifically, is documented.

How to Use It, If You Want To

There is nothing wrong with using rosemary oil as long as it is used as what it is: a low-risk topical layer with weak evidence, not a substitute for treatment in anyone actively losing hair.

Dilute it. One to five percent in a carrier oil — about 3 to 15 drops per 15 ml of jojoba, argan or fractionated coconut oil. Three percent is a sensible standard. Undiluted essential oil on a scalp is an irritant, not a stronger dose.

Patch-test it. A 2% dilution on the inner forearm for 48 hours before the first scalp application. If it reddens, itches or blisters, rosemary is not for you.

Apply it once daily — 1 to 2 ml massaged into the thinning areas, left at least 30 minutes or overnight, washed out. Keep it off broken skin and off a freshly microneedled scalp.

Never ingest it. A systematic review of essential oils and seizures lists internal use of rosemary oil — along with sage, camphor, eucalyptus and others rich in 1,8-cineole or camphor — among documented convulsant exposures. There is no oral hair-loss indication for rosemary essential oil and no dose at which swallowing it is a good idea.

Photograph your scalp under the same light, at the same distance, monthly. The trial saw nothing at three months. If nothing has happened by six months of standardised photographs, it has not worked, and the decision moves up a tier.

Why It Does Not Replace Minoxidil

The argument is not that rosemary is useless. It is that androgenetic alopecia is progressive, and follicles that fully miniaturise do not recover. Placebo arms in the finasteride trials lost hair steadily over two years; that is what untreated pattern loss does.

For a person whose hair is actively thinning, six months on an agent whose best-case ceiling is "equivalent to the weaker concentration of the second-line drug" — and whose realistic case is unknown because the trial had no control — is six months of follicles that a 5% minoxidil solution, or a 5-alpha-reductase inhibitor, might have kept. That is the cost of the substitution, and it is not refundable.

For a person with early, slow thinning who has declined minoxidil with open eyes, rosemary is a defensible Layer 1. For a person already on minoxidil, it is at most an adjunct applied at a different time of day. Where it sits, and what should come before and after it, is set out in The Hair Retention Protocol. The biology of why early treatment matters more than the choice of agent is in Hair Loss Science: DHT, Follicles and What Actually Works.

Frequently Asked Questions

Did the trial show rosemary oil is as effective as minoxidil?

It showed no significant difference between rosemary oil and minoxidil 2% over six months, in a trial with no placebo arm. Whether that means both worked equally or neither worked much cannot be determined from the design. And 2% is not the standard concentration — 5% gave 45% more regrowth in the pivotal comparison.

Why does the missing placebo arm matter so much?

Because hair counts drift. Patients enrol at a low point, counts vary with season and shedding cycle, and a similar six-month rise in two arms is exactly what a null result looks like without a control. A placebo arm would have shown whether the rise exceeded that background drift. Its absence leaves the question open.

What about the mouse and test-tube studies?

They used rosemary leaf extract, not the essential oil, at concentrations that describe a dish rather than a scalp. The 5-alpha-reductase inhibition (82–95%) and the mouse regrowth are real findings about the extract's constituents; they do not establish that a diluted oil delivers an active dose to a human follicle.

Can I use rosemary oil together with minoxidil?

There is no interaction data either way. If a clinician has prescribed minoxidil, it stays; rosemary can be applied at a different time of day as an adjunct. Do not mix the two in one formulation, and do not treat rosemary as a reason to taper the drug.

Is rosemary oil safe?

Topically, diluted, on intact skin, in someone who has patch-tested — yes, with the documented exception of allergic contact dermatitis. Undiluted it is an irritant. Ingested, rosemary essential oil appears on the list of convulsant essential oils. Never swallow it.

How long before I know if it is doing anything?

The trial saw nothing at three months in either arm. Judge at six months of standardised, same-light photographs. If there is no visible change by then, escalate rather than extend.

Related Research

• Rosemary Oil (Topical)
• The Hair Retention Protocol
• Hair Loss Science: DHT, Follicles and What Actually Works
• The Hair Retention Protocol: Topical and Systemic Support Built in Evidence Order
• GHK-Cu for Skin and Hair