Best hair loss treatments, ranked by evidence.

Ranked on: the strength and consistency of human randomised controlled trial evidence for slowing or reversing hereditary hair loss. Mechanism, animal data and seller photographs never lift a grade. No commercial relationship influences the order.

1. 01

   Topical minoxidil 5%

   GRADE A·EstablishedThe best-evidenced treatment you can buy without a prescription

   Large, repeated, placebo-controlled trials in both men and women show increased hair count and density over 16 to 48 weeks. It works by opening ATP-sensitive potassium channels and prolonging the follicle's growth (anagen) phase, which both thickens miniaturised hairs and slows further loss. Foam is usually better tolerated than solution because it omits propylene glycol, the usual cause of scalp itch and irritation. Two caveats matter: a burst of shedding in the first two to eight weeks is expected and is a sign of follicles resetting into a new growth cycle, not a reason to stop; and the benefit persists only while you keep applying it.

2. 02

   Oral finasteride 1 mg

   GRADE A·EstablishedStrongest evidence for stopping loss, prescription only

   Finasteride inhibits type II 5-alpha-reductase, the enzyme that converts testosterone to dihydrotestosterone, and it is DHT that drives follicular miniaturisation in genetically susceptible scalps. Multi-year randomised trials in men show most participants stop losing hair and a substantial minority regrow, with effects strongest at the crown. It is the closest thing to a disease-modifying treatment in this list because it acts on the cause rather than the growth cycle. It is not approved for female pattern hair loss in the same way, is contraindicated in pregnancy because of the risk of genital malformation in a male fetus, and a minority of men report sexual side effects — discuss this with a prescriber before starting, not after.

3. 03

   Dutasteride 0.5 mg (off-label in most markets)

   GRADE B·PromisingMore potent than finasteride, thinner long-term safety record

   Dutasteride inhibits both type I and type II 5-alpha-reductase and suppresses scalp DHT more completely than finasteride. Head-to-head randomised trials show greater hair count gains, and it is licensed for hereditary hair loss in South Korea and Japan while remaining off-label for that use elsewhere. The trade-offs are a long elimination half-life measured in weeks rather than hours and a smaller long-term safety database in this specific indication, which is why most prescribers reserve it for people who have plateaued on finasteride rather than using it first line.

4. 04

   Low-dose oral minoxidil (0.25–5 mg)

   GRADE B·PromisingFast-growing evidence, real cardiovascular monitoring needed

   Originally an antihypertensive, taken at a fraction of the blood-pressure dose it has become a mainstream dermatology option, with randomised comparisons against topical minoxidil now supporting it and a large body of retrospective clinic data behind that. Its practical advantage is compliance: a tablet beats twice-daily scalp application for most people, and it reaches follicles that a topical misses. Its costs are systemic — unwanted body and facial hair is common and dose-dependent, and fluid retention, ankle swelling and palpitations occur, so it requires a prescriber who checks blood pressure and screens for cardiac disease rather than a telehealth checkbox.

5. 05

   Low-level laser therapy (caps, helmets, combs)

   GRADE B·PromisingModest but genuine sham-controlled evidence

   Several randomised, sham-device-controlled trials of red-light devices around 650–680 nm show statistically significant increases in hair count over 16 to 26 weeks. The honest framing is that the effect size is small next to minoxidil or finasteride, the trials are mostly industry-funded and short, and the devices cost several hundred dollars. It is a reasonable add-on for someone already on the two first-line treatments who wants more, and a poor substitute for either of them.

6. 06

   Ketoconazole 2% shampoo

   GRADE C·EarlyThin but real data; cheap enough to be worth it

   Small controlled studies report improvements in hair density and shaft diameter with ketoconazole shampoo, plausibly through reducing scalp inflammation and Malassezia colonisation, with a weak local anti-androgenic effect also proposed. The evidence base is a handful of small studies rather than a body of trials, and shampoo contact time is measured in minutes, which caps how much any active can do. Two or three washes a week alongside the treatments above is a sensible, low-cost adjunct — not a treatment in its own right.

7. 07

   Microneedling (dermaroller or dermapen)

   GRADE C·EarlySmall trials, mostly as a minoxidil amplifier

   Randomised trials exist, but they are small, single-centre and heterogeneous, and almost all test microneedling plus minoxidil against minoxidil alone rather than microneedling on its own. Those trials do favour the combination, which is the reason to take it seriously. Needle depth, frequency and device quality vary wildly between studies and between clinics, there is no standard protocol, and home rolling at aggressive depths risks infection and scarring. Treat it as an optional amplifier of a working regimen, weekly at most, on an intact scalp.

8. 08

   Platelet-rich plasma (PRP) injections

   GRADE C·EarlyHeterogeneous small trials, no standard preparation

   PRP has a genuine signal in several small randomised and split-scalp studies, but preparation protocols — centrifuge speed, platelet concentration, activation, injection depth, session interval — differ so much between studies that the results cannot be pooled into a dependable effect size. In practice that means the PRP you buy is not necessarily the PRP that was studied. It is expensive, needs repeated sessions indefinitely, and belongs after minoxidil and a 5-alpha-reductase inhibitor have been given a fair trial, never instead of them.

9. 09

   Topical peptide serums and rosemary oil

   GRADE D·Insufficient or unsafeWeak evidence; marketed far beyond it

   Copper-peptide and 'growth factor' serums are sold on laboratory and cosmetic-appearance data, not on hair-count trials in hereditary hair loss. Rosemary oil rests essentially on one small trial comparing it to 2% minoxidil with no placebo arm — a comparison against a weaker dose of an active drug, which cannot establish that either arm beat doing nothing. Neither is dangerous, and neither belongs in the budget ahead of 5% minoxidil, which costs less and has hundreds of times the evidence.

10. 10

    Biotin, collagen and 'hair, skin and nails' supplements

    GRADE D·Insufficient or unsafeNo evidence unless you are genuinely deficient

    Biotin corrects hair loss caused by biotin deficiency, which is rare outside specific genetic disorders, long-term anticonvulsant use or prolonged malabsorption. In people with normal levels there is no randomised evidence that it grows hair, and high-dose biotin has a concrete harm: it interferes with streptavidin-biotin laboratory immunoassays, skewing troponin, thyroid and hormone results badly enough to have caused missed heart attacks. Collagen and multi-ingredient gummies have no hair-count trials in hereditary loss at all. Worth testing and correcting iron, ferritin, vitamin D and thyroid function if hair is shedding diffusely — not worth buying blind.

Reading a hair loss shampoo label honestly

• Ketoconazole 2% is the only shampoo active with controlled hair-density data behind it, thin as that data is — 1% versions are widely sold and less studied.
• Pyrithione zinc and salicylic acid treat flaking, scale and scalp inflammation; that is worth having, but it is scalp care, not hair growth.
• Caffeine, saw palmetto, biotin and 'DHT-blocking' botanicals in shampoo have no dependable hair-count trials — contact time of two minutes is the structural problem no formula solves.
• Ignore 'clinically proven' on a bottle unless the claim names the active, the concentration and a controlled trial; consumer-perception surveys are routinely marketed as clinical proof.
• Thickening, volumising and fibre-coating products make existing hair look denser and change nothing about how much of it there is. That is a cosmetic purchase, and a legitimate one — just do not count it as treatment.

The practical notes most hair loss advice skips

• Nothing in this list shows visible change before four months, and six is the fair judging point. Most people who conclude a treatment failed stopped at week ten, during the expected shedding phase.
• Finasteride is contraindicated in pregnancy and women who are or may become pregnant should not handle crushed or broken tablets; for female pattern hair loss, topical minoxidil is the evidence-based starting point and any anti-androgen is a specialist decision.
• Finasteride lowers PSA by roughly half — tell any clinician ordering a PSA that you take it, or a meaningful result can be read as normal.
• Sudden diffuse shedding across the whole scalp, patches, scaling, scarring or pain is not pattern hair loss. Thyroid disease, iron deficiency, postpartum shedding, alopecia areata and scarring alopecias all need diagnosis before treatment, and some are time-critical.
• Review your own medicine list. Isotretinoin, some antidepressants, beta-blockers, anticoagulants, hormonal contraception changes and high-dose vitamin A can all drive shedding, and that is fixable in a way genetics is not.
• Adding four things in one month guarantees you will not know which one worked. Change one variable at a time, three to six months apart.