Does rosemary oil really work for hair growth?
Rosemary oil performs as well as minoxidil 2% in one small study, but dermatologists do not yet recommend it; feel free to use it as a low-barrier alternative, but do not expect miracles when proven treatments are available.
One randomised trial (100 participants, 6 months) compared rosemary oil with minoxidil 2% in people with hereditary hair loss. After six months, both groups had significantly more hairs, and the difference between the two was not significant. After three months, no effect was visible in either group. That sounds encouraging, but it is only one study, without a placebo group.
Rosemary oil also caused more itching than at baseline in that study, although the itching was less severe than with minoxidil. If you have a sensitive scalp, that is something to take into account.
In animal and laboratory research, rosemary extract inhibits an enzyme that plays a role in hereditary hair loss and stimulates hair growth in mice. That is a plausible explanation for the effect, but results in mice do not automatically apply to humans.
The more positive findings are countered by stronger opposing voices. A Canadian panel of eleven dermatologists placed rosemary oil on the list of 'not recommended' treatments for hereditary hair loss in 2025. A comparative analysis of fifteen treatments ranked rosemary oil lower than dutasteride, minoxidil and finasteride. In short: there is something biologically plausible going on, but the clinical evidence is still very thin.
One systematic review regards rosemary oil as an option for people who object to minoxidil or finasteride, or as a supplement to those treatments. That is a realistic perspective: not as a replacement for proven treatments, but as a relatively safe alternative with limited but non-zero supporting evidence. If you are experiencing serious hair loss, discuss the options with your general practitioner or dermatologist.
All claims are based on: one RCT (n=100, PMID 25842469), one animal/laboratory study (PMID 22517595), one systematic review (PMID 35044013), one network meta-analysis (PMID 41051009), one Delphi consensus guideline (PMID 40986632) and one population study (PMID 39186547). The RCT has no placebo group; the network meta-analysis is limited by the quality of the included studies.