Testosterone Therapy: What It Does, and What It Costs Fertility

No. It does the opposite, and it does so reliably enough that the same mechanism has been developed as a male contraceptive. The AUA Testosterone Deficiency Guideline, Statement 23, reads: "Exogenous testosterone therapy should not be prescribed to men who are currently trying to conceive." The Endocrine Society's clinical practice guideline recommends against starting testosterone therapy in patients who are planning fertility in the near term.

Here is why. Your pituitary reads the testosterone in your blood. Given an outside supply, it stops sending LH, which tells the testes to make their own testosterone, and FSH, which supports sperm production. Sperm production needs a testosterone concentration inside the testis far above what circulates in blood, and that internal concentration comes from LH, not from the gel or the injection.

So the blood test improves and the thing it is standing in for collapses. A man can be told his testosterone is now normal while his sperm count is falling towards zero.

How the two measurements differ, and what a low testosterone result does and does not prove, is set out in testosterone and male fertility: what a low result means.

Do not stop it on your own. Stopping abruptly leaves you with neither your own production nor the replacement, and whoever prescribed it needs to manage the change and the monitoring. Book that appointment, and go in with the timeline in hand so the conversation is about planning rather than reassurance.

The best recovery data comes from pooled individual results of 30 hormonal male contraception studies: 1,549 healthy men, a sperm count every month until they recovered. The median time for sperm concentration to return to 20 million per mL was 3.4 months. Pooled across that cohort, the probability of reaching it was 67% within 6 months, 90% within 12 months, 96% within 16 months, and effectively complete by 24.

Read that honestly in both directions. Recovery was eventually full in essentially everyone -- but those were healthy young men with normal counts before they started, on controlled regimens, monitored monthly. Somebody who has been on testosterone for years, or whose count was low to begin with, is not represented by that curve, and no one has published the equivalent curve for long-term clinical use.

If the androgens were not prescribed -- bought at a gym or online -- the picture is the same mechanism at a larger and unpredictable dose. Across 33 studies and 1,766 users, LH fell by a weighted mean of 3.37 IU/L, FSH by 1.73 IU/L and the men's own testosterone by 10.75 nmol/L during use. Gonadotropins came back to baseline over 13 to 24 weeks after stopping, but testosterone was still below baseline at 16 weeks, and the same review recorded structural and functional sperm changes and reduced testicular volume. Tell the doctor what was actually taken; the recovery plan depends on it.

Freezing sperm is worth asking about at this appointment rather than later. What happens to sperm on androgens is covered in anabolic steroids and sperm damage.

Less than the marketing implies, and more than nothing. The largest placebo-controlled programme is the Testosterone Trials: 790 men aged 65 or older with a testosterone below 275 ng/dL and symptoms, randomised to testosterone gel or placebo gel for a year.

What it found, in the order the evidence supports:

• Sexual function improved. Sexual activity, sexual desire and erectile function were all significantly better on testosterone. The trial's authors called the benefit moderate.
• Mood improved slightly. Men on testosterone reported somewhat better mood and less severe depressive symptoms than those on placebo.
• Vitality did not improve. On the FACIT-Fatigue scale, the pre-specified measure of the thing most men actually come in complaining of, there was no significant benefit.
• Walking distance did not improve in the trial designed to test it. The Physical Function Trial's own primary outcome was not significantly different between groups. The often-quoted figure -- 20.5% on testosterone versus 12.6% on placebo achieving a 50-metre improvement in 6-minute walking distance -- reached significance only when men from all three trials were pooled together.
• Memory and cognition did not improve. A separate trial within the same programme randomised 493 men who had both low testosterone and age-associated memory impairment, the group most likely to benefit. The difference in delayed paragraph recall after a year, across those 493 randomised men, was -0.07 (95% CI -0.92 to 0.79), P = .88, with no effect on visual memory, executive function or spatial ability.

Those trials were run in men aged 65 and over, which is not the man reading this page if he is trying to conceive. That is a real limitation and it cuts both ways: the benefits shown there may be larger or smaller in a 34-year-old, and nobody has run the equivalent trial in one.

This is where most writing on testosterone is a decade out of date, in both directions.

Cardiovascular risk was the big open question and it has now been tested directly. TRAVERSE randomised 5,246 men aged 45 to 80 who had pre-existing or high risk of cardiovascular disease and two fasting testosterone values below 300 ng/dL, and followed them for a mean of 33 months. The primary composite of cardiovascular death, non-fatal heart attack and non-fatal stroke occurred in 7.0% on testosterone versus 7.3% on placebo, hazard ratio 0.96 (95% CI 0.78 to 1.17), meeting non-inferiority. That is not a licence -- it is the absence of the specific harm that had been feared.

The same trial found three things that were more frequent on testosterone and that get left out of the summaries: atrial fibrillation, acute kidney injury and pulmonary embolism. Those are the risks worth asking your prescriber about.

Prostate cancer has also been measured rather than assumed. Pooling the prospectively collected prostate data from four randomised trials including TRAVERSE, in men screened to exclude those at high prostate cancer risk, the rates of any prostate cancer, high-grade prostate cancer, acute urinary retention, surgery for benign prostatic hyperplasia, prostate biopsy and new drug treatment for urinary symptoms were low and did not differ from placebo. Testosterone did not worsen urinary symptoms. It did raise PSA more than placebo in the first year, which is why a PSA and prostate assessment before starting, and a monitoring schedule after, are part of the guideline rather than an optional extra.

The Endocrine Society's list of situations in which it recommends against starting testosterone at all is worth reading before the first prescription: planning fertility in the near term, breast or prostate cancer, a palpable prostate nodule or induration, PSA above 4 ng/mL, elevated haematocrit, untreated severe obstructive sleep apnoea, severe lower urinary tract symptoms, uncontrolled heart failure, a heart attack or stroke in the last 6 months, and thrombophilia.

They are permitted. That is not the same as proven, and the difference matters if you are being asked to pay for one.

The idea is sound: instead of supplying testosterone from outside, raise the signals above it. Clomiphene and other selective oestrogen receptor modulators, aromatase inhibitors such as letrozole and anastrozole, and injected hCG all push LH or mimic it, so testosterone rises without the axis being switched off. The AUA guideline's Statement 27 says clinicians "may" use aromatase inhibitors, hCG, selective oestrogen receptor modulators or a combination in men with testosterone deficiency who want to maintain fertility -- conditional language, chosen deliberately.

The reason it is conditional: a 2026 systematic review in the Journal of Clinical Endocrinology and Metabolism searched to February 2025 for hormonal treatment in normogonadotropic men with abnormal semen parameters. Nine studies and 735 men qualified. No eligible study evaluated clomiphene. None evaluated an aromatase inhibitor. Tamoxifen showed no benefit and neither did combined hCG with hMG. Only injected FSH produced measurable semen changes, which the reviewers called modest and short-term with uncertain translation to pregnancy or live birth.

One group is different, and it is worth knowing whether you are in it. Where gonadotropin deficiency is the proven cause -- the pituitary genuinely not sending the signal -- gonadotropin treatment does induce sperm production. Across 41 studies, after a median 18 months of hCG or hCG with FSH, sperm appeared in the ejaculate in 78% of 1,673 pooled patients but exceeded 10 million per mL in only 24%. That is a real treatment with a modest ceiling, not a general answer for low testosterone.

One historical note, because it still props up confident writing on this: Cochrane withdrew its review "Clomiphene or tamoxifen for idiopathic oligo/asthenospermia" in 2007, and the record is titled "WITHDRAWN". It should not be cited as current evidence. Clomid for men: what the evidence does and does not show goes through the clomiphene trials, and hormonal therapy for male infertility covers the injected options.

Four things, and they are all cheap relative to what they prevent.

1. Whether the diagnosis is confirmed. The AUA guideline requires two total testosterone measurements on separate occasions, both taken in the early morning, and uses a total below 300 ng/dL as a reasonable cut-off. The Endocrine Society adds that hypogonadism should be diagnosed only in men who have symptoms and signs as well as consistently low testosterone. A single afternoon sample in a man who feels well is not a diagnosis.
2. Whether you might want children, and on what timescale. This is the question that changes the answer, and it is the one most often not asked.
3. A baseline semen analysis, and a conversation about freezing sperm. If you start testosterone without one, you will not know later whether a poor count is the drug or was always there.
4. What the monitoring schedule is. The guideline expects symptoms, adverse effects, adherence, serum testosterone and haematocrit to be checked on a standard plan, with prostate cancer risk evaluated during the first year.

Then there is the slower route that is often the right one. Weight, untreated sleep apnoea, opioids, corticosteroids and poorly controlled diabetes all lower testosterone, and treating those raises it without anyone prescribing a hormone. It is less satisfying than a prescription and it does not cost your fertility.

What is and is not worth trying on that route is covered in how to naturally boost testosterone and male fertility and weight.

Being told there is a treatment for how you feel, and then being told to wait for it until after a pregnancy that may take years, is a genuinely hard thing to be told. It is worth saying that the sequencing is the whole of the problem here, not the treatment itself.

• That any dose or route of testosterone spares sperm production. No source retrieved for this article shows a low-dose, microdosed or transdermal regimen that avoids suppressing spermatogenesis.
• That recovery figures apply to long-term clinical use. They come from 1,549 healthy men on monitored contraceptive regimens with normal baseline counts. The equivalent data for years of prescribed testosterone has not been published.
• That clomiphene, letrozole or anastrozole improve a reproductive outcome in a man whose LH and FSH are already normal. The 2026 review found no qualifying trial of any of them.
• That testosterone improves memory, cognition or general vitality. Both were pre-specified, measured in randomised trials, and not found.
• That testosterone causes prostate cancer. Measured across four randomised trials in screened men and not found -- which is not the same as established safety in men at high prostate cancer risk, who were excluded from those trials.
• That testosterone is safe for the heart in men who were excluded from TRAVERSE, including those with a heart attack or stroke in the preceding 6 months, uncontrolled heart failure or thrombophilia. They were not studied.
• That any supplement raises testosterone enough to change a symptom or a fertility outcome. No source retrieved here tested that.