Is Scalp DHT Really Maxed Out on Finasteride, or Can You Push It Lower?

Short answer: The DHT number on a standard blood test is not the number that drives hair loss. In the one trial that measured both compartments in the same men, finasteride 5 mg dropped serum DHT about 73% but scalp DHT only about 41%, and dutasteride 0.5 mg cleared serum DHT about 92% while leaving roughly half of scalp DHT intact (Olsen 2006). The two drugs fall short for different reasons, and that distinction is the useful part. Finasteride is limited by coverage: a hundred-fold dose range, from 0.05 mg to 5 mg, changes scalp DHT by only about eight percentage points, so more drug cannot reach the pool it is missing (Drake 1999). Dutasteride at the standard dose is limited by exposure: going from 0.5 mg to 2.5 mg moved scalp DHT from 51% to 79% while serum barely shifted, which means the enzyme in the scalp was not saturated. Neither limit is visible on a blood test, because serum DHT is already near its floor in every one of these arms. Scalp DHT still cannot be driven to zero with today's drugs, and the regrowth payoff runs into a ceiling set by biology that has nothing to do with DHT.

Why is scalp DHT the number that actually matters, not serum DHT?

Dihydrotestosterone (DHT) is the androgen that miniaturizes genetically susceptible scalp follicles. The important word there is scalp. The DHT that matters for hair loss is the DHT sitting in the tissue around the follicle, not the DHT floating in your bloodstream.

This distinction is easy to lose, because the DHT people can actually measure is almost always the serum (blood) number. It is what a lab draws, and it is what most patients and clinicians ever see. The unspoken assumption is that serum DHT is a fair stand-in for what is happening at the follicle.

It is not, and there is direct human data showing the gap. The cleanest example comes from a randomized trial that biopsied the scalp and drew blood in the same men, so the two compartments could be compared apples to apples (Olsen 2006). What it found is that the two numbers move by very different amounts on the same drug. A serum DHT that looks crushed on a lab report can sit on top of a scalp compartment that is only partly suppressed.

How far a standard pill actually lowers scalp DHT

Here are the measured, head-to-head values. These are trial numbers from scalp biopsies and matched blood draws, not estimates.

Finasteride 5 mg per day for 24 weeks: serum DHT down about 73%, scalp DHT down only about 41% (Olsen 2006). Dutasteride 0.5 mg per day for 24 weeks: serum DHT down about 92%, scalp DHT down only about 51% (Olsen 2006). Dutasteride 2.5 mg per day for 24 weeks: serum DHT down about 96%, scalp DHT down about 79% (Olsen 2006).

The pattern is consistent and striking. Every regimen suppressed serum DHT far more deeply than scalp DHT. Dutasteride 0.5 mg nearly abolished DHT in the blood, about 92%, yet cleared only about half of it in the scalp. Serum DHT saturates near its floor long before scalp DHT does.

A fair question at this point is what happens at finasteride 1 mg, since that is the dose almost everyone actually takes for hair. Olsen did not test it. The only trial that measured scalp DHT across a range of finasteride doses is older, and its answer is more interesting than a single number.

What does finasteride 1 mg do to scalp DHT?

Drake and colleagues biopsied 249 men before and after 42 days on placebo or one of five finasteride doses, then measured DHT in the scalp tissue itself (Drake 1999). The full dose series is the most informative thing in this literature, and it is almost never quoted:

Placebo: scalp DHT down 13.0% Finasteride 0.01 mg: down 14.9% Finasteride 0.05 mg: down 61.6% Finasteride 0.2 mg: down 56.5% Finasteride 1 mg: down 64.1% Finasteride 5 mg: down 69.4%

Read the bottom four rows again. Going from 0.05 mg to 5 mg is a hundred-fold increase in dose, and it moves scalp DHT suppression by about eight percentage points. The standard 1 mg hair dose sits in the middle of that plateau at 64.1%.

Finasteride's effect on scalp DHT is essentially maximal at a twentieth of the dose people take. You cannot dose your way past it. That single observation is the strongest evidence that finasteride's residual scalp DHT is not an exposure problem but a coverage problem: there is a pool of DHT in the scalp that finasteride structurally cannot reach, and adding more drug does not help because the drug was never the limiting factor.

Two honest notes on that table. The dose series is not perfectly monotonic, because 0.2 mg scored slightly lower than 0.05 mg, which tells you how much scatter sits in tissue DHT measurement. And the placebo group lost 13.0% of its scalp DHT with no active drug at all, which matters for the next section.

Why do the two trials disagree about finasteride?

Because this is the single most confusing thing in the area, it is worth doing properly. At the same 5 mg dose, Drake measured about 69% scalp suppression and Olsen measured about 41%. Same drug, same dose, a gap of nearly 30 points.

Neither study is wrong. They are not measuring under the same conditions, and four differences are documented in the papers themselves.

The assay is different. Drake extracted androgens from scalp skin and measured them by high pressure liquid chromatography with radioimmunoassay, reporting an interassay precision of 12.1% at 1.7 ng DHT per gram of tissue. With mean baseline levels of 2.0 to 2.4 ng/g, post-treatment samples sit close to that precision floor. Olsen homogenised the tissue and used ether extraction. Notably, Olsen moved serum DHT onto gas chromatography and mass spectrometry specifically because immunoassay could not reliably measure the very low levels that dutasteride produces. Different extraction chemistry and different detection give different recovery, and tissue steroid assays are notoriously method dependent.

The biopsy site is different. Olsen took a 4 mm punch anterolateral to the leading edge of the vertex bald spot, adjacent to the hair count target area. Drake enrolled men who were candidates for hair transplantation or willing to undergo biopsy of balding skin, without the same site specification. This matters more than it sounds, because sebaceous gland density varies across the scalp and sebaceous glands are the type 1 reservoir. Move the punch and you change what proportion of the sample is made of tissue finasteride cannot act on.

The timepoint is different. Drake measured at 42 days, Olsen at 24 weeks. Counterintuitively, the longer study found less suppression, not more.

And the placebo arms were handled differently. Drake's placebo group lost 13.0% of scalp DHT over 42 days with no active drug. Drake's headline figures are raw change from baseline, so that drift is still inside them. Net of placebo, the 5 mg figure is closer to 56% than to 69%, which closes a meaningful part of the distance to Olsen's 41%.

We want to be careful with that last point rather than lean on it. Olsen's placebo scalp DHT change is presented in a figure rather than quoted as a number in the text, so we cannot complete the same adjustment on their side. If their 41% is also raw change from baseline and their placebo drifted similarly, the two are already comparable and the placebo argument does not apply. Nobody has formally reconciled these two trials, and we are decomposing the difference rather than resolving it.

The fair summary is that finasteride's scalp DHT suppression sits somewhere in a roughly 40% to 70% band depending on assay, site and timepoint, and that in every study it is materially less complete than its effect on serum. Finasteride 1 mg is the FDA-approved oral dose for male pattern hair loss; the higher doses and the dutasteride regimens discussed here are used off-label for hair.

Two different ceilings

Putting the two trials side by side gives the clearest picture in this whole area, and it is not the picture most summaries paint.

Finasteride is limited by coverage. A hundred-fold dose range moves scalp DHT by eight points (Drake 1999). More drug does not help, because the leftover DHT is being made by an isoform finasteride barely inhibits.

Dutasteride at the standard dose is limited by exposure. Going from 0.5 mg to 2.5 mg moved scalp DHT from 51% to 79% while serum barely shifted, 92% to 96% (Olsen 2006). More drug helped a great deal, which means the enzyme in the scalp was not saturated at the standard dose even though the compartment setting the blood number already was.

Those are two different problems wearing the same clothes. And neither is visible on a blood test, because serum DHT was already near its floor in every one of these arms.

How do we know what the hair follicle actually sees?

Everything above rests on one measurement, and it is worth being honest about what that measurement is.

A scalp biopsy is a 4 mm punch of whole tissue. It contains epidermis, dermis, blood vessels, immune cells, sebaceous glands and whatever follicles are still there. When a paper reports "scalp DHT," it is reporting the average across all of that.

That matters because of where the two enzyme isoforms sit. Sebaceous glands are large, numerous, and the main home of type 1. They are also full of lipid, and DHT is lipophilic. So it is entirely possible that a bulk punch is mostly reporting a sebaceous pool, while the dermal papilla, the cell that actually drives miniaturisation, is far better suppressed than the tissue average suggests.

If that is right, then the 51% figure for dutasteride understates what the follicle experiences, and the honest reading of this whole area changes.

Has anyone measured DHT closer to the follicle?

One group has tried, and the attempt is instructive both for what it found and for what it cannot show.

Hobo and colleagues measured DHT directly in hair by mass spectrometry, in 1,078 men grouped by whether finasteride or dutasteride was detectable in the sample (Hobo et al 2023). The medians look dramatic. Men with no drug detected sat at 3.05 pg/mg. On finasteride, 0.69. On dutasteride, 0.125.

Read quickly, that says dutasteride nearly abolishes DHT at the follicle, which would settle the argument. Three details stop it doing that.

It is hair, not follicle. The method takes ten strands and analyses a 3 cm segment measured from the root, explicitly as a three-month integrator. A terminal anagen follicle is only about 3 to 5 mm deep, so roughly nine tenths of what was analysed is shaft that had already emerged from the skin. The dutasteride figure is mostly imputed. The assay's lower limit of quantification was 0.25 pg/mg, and only 11 of 62 dutasteride samples cleared it. Values below were substituted with 0.125, half the limit. The group's median and both quartiles are that substituted number. "Virtually eliminated" is more precisely "below the floor of the assay in 51 of 62 men." No blood DHT was collected. The authors list this among their limitations. It matters because the same group's method paper describes hair steroid levels as reflecting "the integrated values of hormone secretion over the past few months," and lists detection of banned drug use among the applications. That is a description of systemic exposure. Dutasteride lowers serum DHT by 92% to 96%, so a large fall in hair DHT is exactly what circulation alone would predict.

Does the anatomy help settle it?

Partly, and in a direction that is easy to get backwards.

The hair matrix is epithelium, and epithelium has no blood supply of its own. There is no vessel running to the matrix. Everything it receives arrives by diffusion across the basement membrane, and the nearest source is the capillary loop inside the dermal papilla, with a perifollicular plexus supplying the rest of the follicle. Recent imaging work found the vessels around the papilla are the ones preferentially mobilised as follicles grow (Zeng et al 2026).

So there is no route by which a hormone is built into the hair shaft while bypassing the papilla's neighbourhood. Whatever ends up in the fibre diffused through the bulb on the way in. Hair DHT is therefore not simply a blood measurement wearing a disguise.

But it is still not a measurement of the thing that causes hair loss. What drives miniaturisation is DHT bound to androgen receptors inside dermal papilla cells, much of it made in place. Locally synthesised DHT can occupy a receptor inside the cell that made it without ever joining the diffusible pool that gets swept into the keratinising shaft. The fibre samples the extracellular pool. The pathology runs through the intracellular one. A large fall in the first is compatible with meaningful occupancy remaining in the second.

One tempting idea can be set aside on anatomy alone. Sebaceous DHT is unlikely to reach the papilla in a terminal follicle: the sebaceous duct empties into the follicular canal at the isthmus, flow from there is outward toward the skin surface, and below that point the inner root sheath tightly ensheathes the shaft, leaving no lumen to travel down. This is inference from structure rather than a measurement anyone has published, and it may not hold in miniaturised follicles, where the follicle shortens dramatically while the sebaceous gland stays relatively large.

What would actually settle this?

Measuring DHT in microdissected dermal papilla and in whole scalp biopsy, from the same men, on the same drug.

Nobody has done it. Until someone does, every number in this article is a tissue average that includes compartments the follicle does not care about, and the hair-based alternative is a fibre average that cannot separate local suppression from less hormone arriving in the blood.

That is not a reason to discard the biopsy data. It is the best direct evidence available and the compartment comparison it enables is real. It is a reason to hold the conclusions at the right width: serum DHT clearly overstates suppression at the scalp, the scalp compartment is clearly not maximally suppressed at standard doses, and how much of the leftover is sitting in tissue that matters for hair is genuinely unknown.

Why there is always DHT left in the scalp

The residual is not a delivery accident. It is built into the anatomy of the enzyme that makes DHT.

DHT is produced by 5-alpha-reductase, and in skin that enzyme exists as two relevant versions in two different places:

Type 1 lives mainly in the sebaceous (oil) glands (Thiboutot 2000). Type 2 is found in the hair follicle, including the root sheaths and infundibulum, in scalp-specific mapping (Bayne 1999); type 1 is nonetheless the predominant isoform across scalp skin overall.

In a balding scalp, both isoforms are elevated compared with a non-balding scalp, along with the androgen receptor, while aromatase (which diverts androgens away) runs lower in the balding region (Sawaya and Price 1997). So the balding scalp is a two-enzyme environment, not a pure follicle-only one.

That is the key to the residual. Finasteride is a potent inhibitor of type 2 but only a weak inhibitor of type 1. It shuts down the follicle enzyme well, but it largely cannot touch the type 1 enzyme in the sebaceous glands, which keeps generating DHT locally. That sebaceous, type 1 pool is a mechanistic reservoir of leftover scalp DHT that a type-2-selective drug structurally cannot reach.

A few other contributors add to the floor. Oral drug reaches the follicular and sebaceous microenvironment imperfectly, so local enzyme is never fully saturated. Adrenal precursors and local steroidogenesis feed the tissue androgen pool from upstream. And DHT can be made through a so-called backdoor route that partly bypasses the exact testosterone-to-DHT step the drug blocks (a mechanistic, modeled contributor rather than one quantified in human scalp). The net effect is that the scalp keeps its own small DHT supply running even when the blood is nearly dry.

What else the research shows can lower scalp DHT

The research points to two things that lower scalp DHT beyond what a standard finasteride pill achieves, each with a systemic cost that a clinician has to weigh. This is a description of what the trials measured, not a protocol; every regimen below is off-label to some degree and is a decision to make with a doctor, not a knob to turn on your own.

The first is inhibiting both enzyme isoforms instead of one, which is covered above. The important caveat belongs here rather than in the mechanism section: the arm that reached 79% scalp suppression was dutasteride at 2.5 mg, five times the 0.5 mg dose used off-label for hair, studied for only 24 weeks. Systemic DHT suppression rises with dose, and that dose-dependent profile is exactly why standard care stops at 0.5 mg rather than chasing the last fraction of scalp DHT. The 2.5 mg arm is useful as mechanistic proof that residual scalp DHT exists. It is not a destination.

The second is topical delivery, which is compartment-selective. The logic is to put the drug into scalp skin to suppress DHT locally while limiting how much reaches the bloodstream. A topical finasteride 0.25% solution lowered scalp DHT about 47% to 52% while cutting serum DHT only about 24% to 26% at low application volumes (Caserini 2016), and a manufacturer-run phase III trial of a topical finasteride spray reported hair-count results in the range of oral finasteride with a smaller serum DHT hit (about 35% versus about 56%) (Piraccini 2022). That was a single sponsor-run trial framed around non-inferiority, so read "in the range of oral" as that trial's finding rather than settled equivalence.

There is an important nuance here that gets oversold. Topical finasteride's scalp DHT reduction (about 47% to 52%) sits in the same band as oral finasteride. Its real advantage is a better ratio of scalp suppression to serum suppression, not a deeper scalp floor. And the incremental scalp-DHT drop from stacking a topical on top of an oral dual inhibitor has never been measured in a trial, so any claim that a particular stack "maximizes" scalp DHT is modeled, not measured. The reachable scalp floor with current pharmacology appears to bottom out around 20% to 25% residual (roughly 75% to 80% suppression) even under aggressive dual inhibition, and getting there costs disproportionately more systemic exposure (a value extrapolated from the trial ceiling, so modeled rather than directly measured).

Does pushing scalp DHT lower actually grow more hair?

This is the core of the question, and the answer has two halves that are both true.

Over the short term, deeper scalp suppression did track with more hair. In the same 24-week trial, dutasteride 2.5 mg (scalp DHT about 79% suppressed) produced greater hair counts than dutasteride 0.5 mg (about 51%) and finasteride 5 mg (about 41%) (Olsen 2006). Because serum DHT was already floored across all three arms, the growth ranking followed the scalp number, not the serum number. That is a strong measured signal that scalp DHT is the operative variable and that there is regrowth headroom above the standard regimen, at least over 24 weeks. It is worth keeping the same caution in view: the 2.5 mg dose that produced that signal is five times the off-label hair dose and was tested only for 24 weeks, so the signal is a reason to take scalp DHT seriously, not a reason to escalate dose.

But that does not mean lower is always better without limit. No long-term, multi-year trial has shown that maximizing scalp DHT beyond the standard dutasteride 0.5 mg regimen produces durably more hair. The 2.5 mg advantage is a 24-week, surrogate-and-count signal, not a years-long outcome. And efficacy is bounded by biology that DHT removal cannot fix: AGA is polygenic, involves androgen receptor density and local signaling, and drives a follicular miniaturization that becomes progressively less reversible. Even a hypothetical 100% scalp DHT suppression would not be expected to regrow follicles that have regressed past the point of response. Long-term clinical data show maintenance more than unlimited continued gain, which is exactly what a bounded dose-response looks like.

So both statements hold, at different points on the curve. Going from partial (finasteride-level) to deeper (dual-inhibitor-level) scalp suppression buys measurable extra growth in the short term, the ascending part of the curve. Then the curve flattens: the last fraction of scalp DHT is the hardest to remove, the follicles still able to respond get fewer, and non-DHT biology caps the result.

What is the takeaway for reading DHT numbers?

Two habits do most of the work. First, when you see a DHT suppression figure, ask which compartment it describes. A serum DHT number, the kind on a lab report, systematically overstates how suppressed the follicle actually is. Second, ask which enzyme isoform the drug covers. A type-2-selective drug leaves a type-1 sebaceous reservoir untouched by design, which is most of why a "crushed" blood number can coexist with a scalp that is only half cleared.

At Anagen, that is the discipline we bring to every androgen number we cite: not "how low did the blood go," but "what saw it, and is that the compartment that actually matters."

Frequently asked questions

Does a low serum DHT on my blood test mean my scalp DHT is low too?

Not necessarily. In a head-to-head trial, finasteride 5 mg cut serum DHT about 73% but scalp DHT only about 41%, and dutasteride 0.5 mg cut serum about 92% while leaving roughly half of scalp DHT (Olsen 2006). The blood number consistently overstates how suppressed the follicle is.

How much does finasteride 1 mg lower scalp DHT?

About 64% at 42 days, measured directly in scalp tissue (Drake 1999). The more useful finding from that trial is how flat the dose response is: 0.05 mg produced about 62% and 5 mg produced about 69%, so a hundred-fold increase in dose changes scalp DHT by roughly eight percentage points. Finasteride's effect on scalp DHT is close to maximal at a small fraction of the standard hair dose.

Why do studies disagree about how much finasteride lowers scalp DHT?

At the same 5 mg dose, one trial measured about 69% suppression at six weeks and another about 41% at 24 weeks. The documented differences are the assay (HPLC with radioimmunoassay versus homogenisation and ether extraction), the biopsy site (one specified a 4 mm punch anterolateral to the leading edge of the vertex bald spot, the other did not), the timepoint, and the handling of placebo drift, since the placebo group in the earlier trial lost 13% of scalp DHT with no active drug and that drift is still inside its headline figures. Both are legitimate measured values. The honest band is roughly 40% to 70%.

Does a scalp biopsy measure DHT in the hair follicle?

Not specifically. A 4 mm punch is whole tissue: epidermis, dermis, blood vessels, immune cells, sebaceous glands and follicles. Reported scalp DHT is the average across all of it. Because sebaceous glands are large, numerous, lipid-rich and the main site of type 1 5-alpha-reductase, a bulk measurement may substantially reflect a sebaceous pool rather than the dermal papilla that actually drives miniaturisation.

Has anyone measured DHT in hair follicles directly?

Not in the dermal papilla. One group measured DHT in hair itself by mass spectrometry and found much lower levels in men on dutasteride (Hobo et al 2023), but that analysis used a 3 cm segment of hair shaft as a three-month integrator, most of which had already grown out of the skin, and 51 of 62 dutasteride samples fell below the assay's limit of quantification. No blood DHT was collected, so local suppression cannot be separated from less hormone arriving via circulation. Measuring microdissected dermal papilla against whole biopsy in the same men has not been done.

Which of these drugs is FDA-approved for hair loss?

Oral finasteride 1 mg is FDA-approved for male pattern hair loss. Dutasteride, the higher finasteride doses, and topical finasteride or dutasteride are used off-label or as compounded or region-specific products. Any use should be decided with a qualified clinician.

References

1. Olsen EA, Hordinsky M, Whiting D, et al. The importance of dual 5alpha-reductase inhibition in the treatment of male pattern hair loss: results of a randomized placebo-controlled study of dutasteride versus finasteride. J Am Acad Dermatol. 2006;55(6):1014-1023. DOI: 10.1016/j.jaad.2006.05.007. PMID: 17110217.
2. Drake L, Hordinsky M, Fiedler V, et al. The effects of finasteride on scalp skin and serum androgen levels in men with androgenetic alopecia. J Am Acad Dermatol. 1999;41(4):550-554. DOI: 10.1016/S0190-9622(99)80051-6. PMID: 10495374.
3. Thiboutot D, Bayne E, Thorne J, et al. Immunolocalization of 5alpha-reductase isozymes in acne lesions and normal skin. Arch Dermatol. 2000;136(9):1125-1129. DOI: 10.1001/archderm.136.9.1125. PMID: 10987868.
4. Bayne EK, Flanagan J, Einstein M, et al. Immunohistochemical localization of types 1 and 2 5alpha-reductase in human scalp. Br J Dermatol. 1999;141(3):481-491. DOI: 10.1046/j.1365-2133.1999.03042.x. PMID: 10583052.
5. Sawaya ME, Price VH. Different levels of 5alpha-reductase type I and II, aromatase, and androgen receptor in hair follicles of women and men with androgenetic alopecia. J Invest Dermatol. 1997;109(3):296-300. DOI: 10.1111/1523-1747.ep12335779. PMID: 9284093.
6. Caserini M, Radicioni M, Leuratti C, et al. Effects of a novel finasteride 0.25% topical solution on scalp and serum dihydrotestosterone in healthy men with androgenetic alopecia. Int J Clin Pharmacol Ther. 2016;54(1):19-27. DOI: 10.5414/CP202467. PMID: 26636418.
7. Piraccini BM, Blume-Peytavi U, Scarci F, et al. Efficacy and safety of topical finasteride spray solution for male androgenetic alopecia: a phase III, randomized, controlled clinical trial. J Eur Acad Dermatol Venereol. 2022;36(2):286-294. DOI: 10.1111/jdv.17738. PMID: 34634163.
8. Hobo Y, Nishikawa J, Taniguchi Asai N, et al. Evaluation of the therapeutic effects of AGA drugs by measuring finasteride, dutasteride, and dihydrotestosterone in hair. Clin Chim Acta. 2023;547:117456. DOI: 10.1016/j.cca.2023.117456. PMID: 37385468.
9. Hobo Y, Nishikawa J, Miyashiro Y, Fujikata A. Measurement of steroid hormones by liquid chromatography-tandem mass spectrometry with small amounts of hair. Steroids. 2020;164:108732. DOI: 10.1016/j.steroids.2020.108732. PMID: 32976917.
10. Zeng Y, Abe A, Takashima S, et al. Preferential crosstalk between perifollicular capillary vessels and dermal papilla cells. 2026. PMID: 41922417.

---

This article is educational and is not medical advice. It discusses how DHT is measured and suppressed in research, not the diagnosis or treatment of any condition, and it makes no dosing or product recommendation. Several agents discussed are used off-label for hair loss. Talk to a qualified clinician about your own care.