Anabolic Steroids and Hair Loss: Why Finasteride Often Cannot Help

Finasteride blocks an enzyme. Most anabolic steroids never touch that enzyme. That one sentence explains why so many men run finasteride through a cycle, do everything they were told, and watch their hairline go anyway.

I’m not going to spend this article arguing about whether you should use anabolics. But I’m not going to pretend the rest of the risk isn’t there either, so here it is once, plainly: anabolic-androgenic steroids carry documented cardiovascular, hepatic, endocrine and fertility risks, several of the oral compounds named below are 17-alpha-alkylated and hepatotoxic, and in many countries, the United States among them, possession without a prescription is a criminal offence, though the law varies sharply and in the UK personal possession is not an offence even though supply is. None of that is what this article is about, and none of it stops being true because this article is about something else.

What this is about is a specific pharmacological mistake that costs people their hair, and it’s a mistake almost everyone in the gym is making at once, because the advice circulating is one generation behind the pharmacology.

How much hair loss are we talking about

Less than the forums suggest, and the truthful answer is that nobody really knows, because the evidence is survey-grade.

The best prospective data comes from a one-year cohort that followed 100 men through a cycle. Self-reported hair loss went from 2% at baseline to 12% by the end of the cycle, roughly three and a half months.^[1] A six-fold rise. Cross-sectional surveys put it higher: 14.2% among Saudi gym users,^[2] 23.7% among Brazilian competitive bodybuilders.^[3] A Danish registry of 1,189 users recorded fewer than five individuals reporting alopecia at all.^[4]

Two things follow. Hair loss on cycle is real and common enough to plan around. And the range from under 0.4% to 23.7% tells you the underlying research is asking men to remember, not measuring their heads.

The mechanism, and where the advice goes wrong

Testosterone isn’t what miniaturises a genetically susceptible follicle. Dihydrotestosterone is. Testosterone gets converted to DHT by the enzyme 5-alpha-reductase, which sits in high concentration in the scalp, and DHT binds the androgen receptor far more tightly than testosterone does.

Finasteride inhibits that enzyme.^[12] That’s its entire mechanism. It reduces scalp skin DHT by around 64% at 1 mg per day.^[5]

Now here’s the catch, and it’s the whole article. A 5-alpha-reductase inhibitor can only block a conversion that is going to happen. If a compound is already 5-alpha-reduced when you inject it, there is no conversion left to prevent. If a compound is not a substrate for the enzyme at all, the enzyme is irrelevant to it. In both cases finasteride is standing at a door the drug isn’t walking through.

The classification comes from a pharmacology review that states it plainly: several commonly used compounds are either already 5-alpha-reduced, including methenolone, drostanolone, stanozolol, oxandrolone and oxymetholone, or do not undergo significant 5-alpha-reduction in the human body, including boldenone, trenbolone, methandienone, turinabol and fluoxymesterone.^[6]

The nandrolone paradox

This one is counterintuitive and almost nobody gets it right.

Nandrolone is a 5-alpha-reductase substrate, so finasteride does affect it. But 5-alpha-reduction of nandrolone produces dihydronandrolone, and dihydronandrolone binds the androgen receptor less tightly than nandrolone does. Receptor binding work from 1985 states it directly: 5-alpha-reduction increases the affinity of testosterone and decreases the affinity of nandrolone.^[7]

Read that again with a hairline in mind. In tissue rich in 5-alpha-reductase, which the scalp is, nandrolone gets partially deactivated on arrival. That is why it is relatively spared. And it means adding finasteride to a nandrolone cycle blocks the deactivation step, leaving more of the stronger parent compound intact in exactly the tissue you were trying to protect. Mechanistically, finasteride is the wrong drug there. Not neutral. Backwards.

Before you act on that, read this next paragraph, because on its own the point above will get your hair cut off. Almost nobody runs nandrolone by itself. It’s run on a testosterone base, and testosterone is a straightforward 5AR substrate that converts to DHT in the scalp. So on a normal test-plus-deca cycle, finasteride is still blocking the conversion doing most of the follicular damage. The nandrolone interaction argues against expecting finasteride to help with the deca component; it does not argue for dropping finasteride from a cycle that contains testosterone. Those are different conclusions and only the first one follows from the receptor data.

Two more caveats. That receptor work was done in castrated rats and cell lines, not human scalp. And no one has run the human experiment. The mechanism is well established, the clinical consequence is inference.

Trenbolone, and why potency is not the same as pathway

Trenbolone binds the androgen receptor with roughly three times the affinity of testosterone, and it induces less growth in organs that highly express 5-alpha-reductase, because it is metabolised to less potent androgens rather than more potent ones.^[8]

So trenbolone delivers a very heavy androgenic signal directly to the receptor, with no enzymatic amplification step in between. A 5-alpha-reductase inhibitor has nothing to grab hold of. This is the clearest case of a compound where finasteride offers close to zero protection, and where the androgenic pressure on the follicle is nonetheless enormous.

One thing worth flagging, because it undercuts a lot of gym folklore. When androgen receptor binding affinities were measured directly across common compounds, the rank order was methyltrienolone, then nandrolone, then methenolone, then testosterone, then mesterolone. Stanozolol, methandienone and fluoxymesterone all came in below 0.05 relative binding affinity.^[9] The 17-alkylated orals with the fearsome reputations are weak direct receptor binders. The circulating “androgenic ratings” are not from peer-reviewed pharmacology, and I could not find a published source for them. Don’t plan around numbers with no paper behind them.

But will finasteride cost me gains

No, and this one has a proper randomised trial behind it.

139 men were randomised, 102 completed, receiving testosterone with or without dutasteride, which blocks 5-alpha-reductase more completely than finasteride does. At 600 mg per week of testosterone, fat-free mass increased by 7.1 kg with dutasteride and 8.1 kg with placebo. The dose-adjusted difference across all groups was not statistically significant, at p equal to 0.18.^[10]

Twenty weeks, healthy men with suppressed endogenous production, not real-world cycling, and hair was never an endpoint. It still settles the specific question it was asked: blocking DHT does not meaningfully cost you muscle. The “finasteride kills your gains” claim doesn’t survive contact with the trial.

The gap nobody mentions

Here’s the sentence that should be at the top of every article on this subject and never is.

No one has ever tested whether finasteride or minoxidil prevents hair loss in anabolic steroid users.

That isn’t my inference. A 2025 narrative review searched for it, and reports that studies are lacking on the efficacy of available hair loss medications in preventing hair loss among AAS users. Its Table 1, which is the complete world evidence on hair-related side effects in this population, contains four entries: one case report, one prospective cohort, one cross-sectional survey, and one registry study.^[4]

Four entries. That’s the entire literature. The female evidence base is emptier still, with a 2025 review concluding that research has disproportionately focused on male pattern hair loss and male athletes.^[11]

So what does make sense

Reason from the mechanism, because that’s all we have.

If the androgenic pressure is arriving as DHT converted from testosterone, an enzyme inhibitor is the right tool and finasteride does its job. If it is arriving as an already-reduced or non-substrate compound binding the receptor directly, the enzyme is not the bottleneck and the receptor is. That is an argument for acting at the receptor rather than at the enzyme.

Topical antiandrogens work at that level. They compete for the receptor in the scalp rather than blocking a conversion upstream, which means their mechanism does not care whether the androgen arrived pre-reduced. I have written in detail about how RU58841 works and where its evidence actually stands, and about the clinical data on pyrilutamide. Both articles are blunt about the fact that the human evidence is limited. I’m not going to pretend a mechanistic argument is a trial. It’s a mechanistic argument, and on this specific question it’s a better one than the enzyme route.

The other lever is dose and compound selection, which is unglamorous and effective. Testosterone-only at a moderate dose is a very different androgenic load on a follicle than a stack built on trenbolone. If hair matters to you, that choice is upstream of everything else.

Where that leaves you

Find a doctor you can be honest with. Harm reduction is a real clinical discipline and plenty of physicians will monitor bloods without lecturing, which is worth far more than anything in this article.

Take a standardised photo before you start anything, under fixed lighting, and repeat it monthly. Self-report is why the prevalence figures in this field range from under 0.4% to 23.7%, and your memory is no better than theirs.

Know the substrate status of what you are running before you decide what to run alongside it. If everything in your stack sits in the bottom two rows of that table, finasteride is something you’re taking for no mechanistic reason.

Be straight with yourself about the ordering. Hair is downstream of the compound choice. No topical fixes a stack that’s chemically hostile to your follicles, and no one has ever demonstrated that any hair loss drug protects you through a cycle. Anyone who tells you otherwise is extrapolating, including me. The difference is I’m telling you that’s what I’m doing.