Scalp DHT vs Blood DHT: How Much Finasteride and Dutasteride Really Block

A reader wrote to me after a year on finasteride, convinced he had failed the drug. His blood test said DHT was down 70 percent. His hairline said otherwise. He hadn’t failed anything. He had been reading the wrong number.

Every finasteride and dutasteride conversation online runs on one statistic: finasteride lowers DHT by about 70 percent, dutasteride by over 90. Both figures are true. Both are measured in blood. Your hair follicles don’t live in your blood. They live in scalp skin, and scalp skin makes a good share of its own DHT on site, with its own enzymes, which the pills reach to very different degrees.^[5]

The study that measured both numbers side by side exists, it’s twenty years old, and almost nobody quotes the half of it that matters. Let me walk you through it, because once you’ve seen the scalp column you’ll never read a DHT blood test the same way again.

The trial that measured both

In 2006 Elise Olsen and colleagues randomized 416 men aged 21 to 45 to six arms for 24 weeks: dutasteride at 0.05, 0.1, 0.5 or 2.5 mg a day, finasteride at 5 mg a day, or placebo.^[1] Two things make this trial unusual. First, it used finasteride at 5 mg rather than the hair-loss dose of 1 mg, so the finasteride arm is a generous comparator. Second, the investigators took scalp biopsies and measured DHT in the skin itself, not just in serum.

Here is what they found. Finasteride 5 mg cut serum DHT by about 71 percent and scalp DHT by about 41 percent. Dutasteride 0.5 mg, the standard dose, cut serum DHT by about 92 percent and scalp DHT by about 51 percent. Dutasteride 2.5 mg, a dose nobody prescribes, cut serum DHT by 96 percent and scalp DHT by 79 percent.^[1]

Read the orange bars again. The pill that “blocks 92 percent of DHT” blocks roughly half of the DHT where your hair is. The pill that “blocks 70 percent” blocks a bit over a third of it at the scalp. The blood number is real, and it’s the right number if your question is about the prostate or about systemic side effects. If your question is about hair, the scalp number is the one that predicts the outcome, and it’s a much more modest number than the marketing suggests.

For 1 mg finasteride, the dose almost everyone with hair loss takes, the earlier Drake study from 1999 did the same biopsy exercise. Serum DHT fell by about 71 percent and scalp DHT by roughly 64 percent at 1 mg in that trial, which is a better scalp result than Olsen’s 5 mg arm reported.^[2] Different labs, different assays, different years, so I wouldn’t put the two scalp figures on the same chart and compare decimals. The honest summary across both is that finasteride clears somewhere between a third and two thirds of scalp DHT, and dutasteride at the standard dose clears about half. Neither gets near the 90 percent people picture.

Why the scalp is stubborn

DHT is made from testosterone by an enzyme called 5-alpha-reductase, and the enzyme comes in two forms. Type 2 lives in the prostate and in the inner root sheath and dermal papilla of the follicle. Type 1 lives in sebaceous glands, the outer root sheath and general skin.^[5] Finasteride is essentially a type 2 drug. Dutasteride inhibits both.

Blood DHT is dominated by type 2 output, mostly from the prostate, which is why finasteride looks so strong in a blood test. The scalp runs a lot of type 1, and Sawaya and Price showed in 1997 that frontal follicles in balding men carry higher levels of both enzyme types and of the androgen receptor than occipital follicles do.^[5] So the tissue you care about is the tissue with the most local DHT production and the most receptors to catch it. That’s the whole reason dutasteride’s extra reach shows up in hair counts even though its blood number only moves from 71 to 92.

Think of it like heating a house with a radiator in the hallway. Turning the hallway radiator down changes the hallway temperature a lot. The bedroom with its own space heater barely notices. Finasteride turns down the hallway radiator. Dutasteride reaches the space heater as well, but even dutasteride at 0.5 mg leaves half the local heat running, which is a fact worth sitting with before you decide the pill “isn’t working.”

What the scalp number buys in hair

Olsen’s hair counts line up with the biopsy data almost too neatly. Over 24 weeks, in a 2.54 cm target circle, the placebo group lost about 32 hairs. Finasteride 5 mg gained about 72. Dutasteride 0.5 mg gained about 96, and dutasteride 2.5 mg gained about 110.^[1] The arm with 51 percent scalp DHT suppression beat the arm with 41 percent, and the arm with 79 percent beat both.

The larger 2014 trial by Gubelin Harcha, 917 men, this time with finasteride at its real 1 mg dose, found the same ordering: dutasteride 0.5 mg significantly out-grew finasteride 1 mg on hair count and hair width at 24 weeks, with a similar adverse event profile across groups.^[4] And the 2022 network meta-analysis in JAMA Dermatology, pooling 23 studies, put dutasteride 0.5 mg ahead of finasteride 1 mg by about 7 hairs per square centimetre at 24 weeks on total hair count.^[6] Seven hairs per square centimetre over half a year isn’t a different category of result. It’s the difference you’d expect from 51 percent scalp suppression versus something in the forties.

The gap between the two pills is real, and it’s exactly as big as the scalp DHT column says it should be. Not bigger.

Why not just take the 2.5 mg dose?

The 2.5 mg arm is the one people fixate on when they first see this chart: 79 percent scalp suppression, the biggest hair count, so why does the approved dose stop at 0.5? Because the trial wasn’t powered to settle safety at five times the standard dose, and the blood-side data from the urology world explain why nobody has been in a hurry to find out. Clark’s dose-ranging study in 399 men with enlarged prostates found that dutasteride 0.5 mg already suppressed serum DHT by 94.7 percent, with far less person-to-person variability than finasteride’s 70.8 percent, and that 5 mg pushed it to 98.4 percent.^[3] In other words, by 0.5 mg the systemic job is essentially done. Going to 2.5 mg buys you more scalp suppression at the cost of a drug exposure five times higher, with a half-life of around five weeks, for a hair gain over 0.5 mg of maybe 14 hairs in the target circle over six months.^[1]

My reading: the 2.5 mg result is the most useful number in the whole paper and the least useful dose. It’s useful because it proves the point, that scalp DHT suppression and hair growth move together, dose for dose. It’s a poor dose because the same marginal gain is available from adding a topical that acts at the receptor, without multiplying systemic exposure. The arithmetic of the chart argues for combining mechanisms, not for escalating one.

Reading your own bloodwork with this in mind

Men send me their labs constantly, so here’s how I’d interpret the common patterns in light of the scalp column.

Serum DHT down 60 to 75 percent on finasteride 1 mg: normal, expected, says nothing about your hair. Serum DHT “only” down 50 percent: still inside the range the Drake trial saw across individuals, and still consistent with a working drug; variability in serum response on finasteride is large, which is one of the things Clark’s data made explicit.^[2][3] Serum DHT down 90 percent or more on dutasteride: also normal, and also not a promise about the scalp, where the same men are sitting at roughly half suppression. Testosterone up 10 to 20 percent on either drug: expected, because the testosterone that no longer becomes DHT stays as testosterone, and Olsen reported that rise as dose-dependent.^[1]

What no blood test can tell you is the number that decides the outcome. Only a biopsy measures scalp DHT, and nobody takes biopsies to monitor finasteride. So the practical monitoring tool is still the one I keep recommending on this site: standardized photographs, same light, same angle, compared at six-month intervals. The lab result is reassurance that you absorbed the pill. The photograph is the measurement.

What this means if you’re on one of them

Three practical things follow, and I’ll give you my view on each because I spent three years on finasteride before switching.

First, stop using a blood DHT test to judge whether finasteride is “working.” A serum DHT that’s down 70 percent tells you the drug is absorbed and the enzyme is inhibited systemically. It tells you almost nothing about your scalp, and it certainly doesn’t tell you that a 70 percent number should have produced a 70 percent result. Judge the drug by photos at six and twelve months, nothing else.

Second, if you’ve had a modest response to finasteride, the scalp column is the mechanistic reason a switch to dutasteride can add hair. You’re not “resistant.” You’re carrying a type 1 enzyme load the drug was never built to touch. Whether that trade is worth it depends on side effect tolerance, cost and your doctor, and dutasteride’s longer half-life changes the rhythm of stopping, which I cover in the switching guide.

Third, even the strongest pill leaves local DHT on the table, and that’s where the receptor-level treatments come in. A topical antiandrogen doesn’t care how much DHT is made. It competes at the receptor, which is why I think the pill plus a scalp-level blocker is the logical pairing, and why the pipeline of receptor antagonists and degraders matters so much. The pill handles production. Something else needs to handle the receptor.

The one number to remember

Fifty-one. That’s the percentage of scalp DHT dutasteride removes at the dose you’d take. Finasteride sits somewhere below it. Everything else you’ll read, the 70s and the 90s, is a blood figure describing a different organ. The follicle reads the local number, and the local number is modest, which is the single best explanation I know for why these drugs stabilize most men and transform only a few.