What's actually happening

On July 15, 2026, Defense Secretary Pete Hegseth posted a video announcing that service members age 30 and older will get their testosterone checked every year as part of their regular health assessment. Troops under 30 can ask for the test if they want it. If a doctor recommends treatment, taking it is the service member's choice. A Pentagon spokesperson confirmed the screening is mandatory for active-duty and reserve personnel 30 and up.

Hegseth framed it as a matter of keeping troops sharp, saying the goal was making sure they have the right testosterone levels to operate at their best. The post was captioned "The High-T Department of War."

It is a reasonable-sounding idea. Testosterone is real. It does real things. Soldiers need to be strong. So test it, fix it, done.

Except the science here is much stranger than the slogan. Let's walk through it.

What testosterone actually does

Testosterone gets talked about like it is bottled aggression. It is not. It is a hormone with a boring, important job description.

It helps maintain muscle mass. It keeps bones dense. It tells your body to make red blood cells, which is how oxygen gets around. It plays a role in how you handle blood sugar and fat. Reproduction is on the list too, but it is one item on a long list.

So yes, it matters. And when it is truly low, that shows up in the data.

Low testosterone really is linked to worse health

This part is not controversial.

A large analysis pooling individual data from many long-term studies, covering more than 255,000 participant-years, found that men with testosterone below 7.4 nmol/L (about 213 ng/dL) had a higher risk of dying from any cause. Below 5.3 nmol/L (about 153 ng/dL), cardiovascular deaths went up too.

The UK Biobank, following 149,436 men, found the same general pattern. Men in the lowest fifth of testosterone had about a 14 percent higher risk of death than men in the highest fifth. The European Male Aging Study found that men with severe hypogonadism, meaning genuinely low testosterone plus symptoms, had over five times the risk of death.

Those are serious numbers. Testosterone is a legitimate health signal. Hold that thought, because the next section is where things get interesting.

The plot twist: what actually pushes testosterone down

Here is the finding that reframes the whole conversation.

Everyone assumes testosterone falls steadily with age, like hair volume. A pooled analysis of 21,074 men found that from age 17 to 70, the age-related decline is basically negligible. It only becomes meaningful after 70.

So what is doing the work? Body weight, mostly. The same analysis found that for every one standard deviation increase in BMI, testosterone dropped by 2.42 nmol/L (about 69.7 ng/dL). That is a far bigger effect than age.

The rest of the list, diabetes, high blood pressure, cardiovascular disease, cancer, poor sleep, and low physical activity, are all independently linked to lower testosterone.

Read that list again. It is mostly a list of things that are going wrong elsewhere. Testosterone frequently behaves less like a cause and more like a smoke detector. It is telling you something is burning. Whether the answer is "add more testosterone" or "put out the fire" is the entire debate.

Nobody agrees on what "normal" means

You would think a blood test would give a clean yes or no. It does not.

The Endocrine Society, which writes the guidelines on this, specifically recommends against screening the general population for low testosterone. Their reasoning: there is no validated screening tool, the cost-effectiveness is unclear, and there is not good evidence that treating men who feel fine but test low actually improves anything they would care about.

An actual diagnosis requires at least two separate blood draws, both taken fasting, both in the early morning, both below the normal range, using a reliable lab assay. One number on one afternoon does not qualify.

And personal variation is huge. Your own healthy baseline might sit well above or below the population average. Which means tracking your own number over time tells you more than comparing it to a chart of strangers.

Why soldiers are the worst possible group to test this way

Now the military-specific evidence, which is genuinely fascinating.

Hard training crashes testosterone. Not a little. A lot.

Finnish conscripts had their levels measured before and after 12 days of field training. Testosterone went from 18.2 nmol/L (about 525 ng/dL) down to 7.1 nmol/L (about 205 ng/dL). That is a 61 percent drop, landing them squarely in "clinically low" territory.

Then they rested for three days and it came all the way back.

Australian soldiers in field training with sleep deprivation showed the same suppression and recovered within four days. And among elite US military operators, 43 percent tested below the normal range for their age, which researchers attributed to constant stress exposure rather than any actual disease.

Here is the mechanism, and it is the key to everything. Researchers gave these men a hormone called hCG, which is basically a direct message to the testicles saying "produce testosterone now." The testicles responded normally. The factory was fine. The order had simply stopped coming down from the brain.

In other words, this is not a broken system. It is a working system doing exactly what it is designed to do when you are underslept, underfed, and stressed. The body deprioritizes building tissue when it is busy surviving. Then it switches back on when you eat and sleep.

Which raises an awkward practical question. If you test a soldier the week after a field exercise, you may have measured a temporary and completely appropriate response, then written it down as a disorder.

But does taking testosterone make you a better soldier?

This is the question the whole policy rests on, and the evidence is thinner than you would expect.

In men with confirmed hypogonadism, testosterone replacement therapy does real things. It adds roughly 1.6 kg of fat-free mass, reduces fat mass, and improves some strength measures. The Testosterone Trials, involving 790 men averaging age 72, found modest gains in sexual function, walking distance, bone density, and anemia correction. But it did not consistently improve overall physical function or vitality.

Then there is the one study that tested this exact scenario. Researchers gave physically active men a single 750 mg dose of testosterone undecanoate before a 20-day simulated multi-stressor military operation, with a placebo group for comparison.

The result: testosterone worked, and it didn't.

The treated men held onto lean mass, gaining 0.41 kg while the placebo group lost 1.85 kg. Clear win on the scale and the body scan.

And on performance? Nothing. No improvement in power. No improvement in strength. No improvement in anaerobic capacity. No improvement in aerobic capacity. Every measure of actually doing soldier things declined the same amount in both groups.

That gap keeps showing up across this literature. Testosterone reliably changes what your body is made of. It much less reliably changes what your body can do.

Is it safe? Mostly, with caveats

Cardiovascular risk was the big worry for years. The TRAVERSE trial settled a lot of it.

It enrolled 5,246 men aged 45 to 80 who had low testosterone plus existing or high-risk heart disease, and followed them for an average of 33 months. Testosterone was not worse than placebo for major cardiac events.

But the fine print matters. The testosterone group had somewhat higher rates of blood clots in the lungs (0.9 percent vs 0.5 percent), atrial fibrillation (3.1 percent vs 2.4 percent), and acute kidney injury (2.3 percent vs 1.5 percent).

The FDA responded by removing the black-box warning about major cardiac events from testosterone labels, while adding a new warning about a possible 2 to 4 mm Hg rise in systolic blood pressure.

Worth reading that carefully. "We did not find this specific risk" is not the same sentence as "there are no risks." The TRAVERSE researchers made this point themselves: their trial studied carefully selected, closely monitored patients. It was not an endorsement of handing it out more widely.

The practical problems with screening everyone

Put it all together and a few issues stand out.

Timing is everything, and timing is uncontrollable. A soldier's testosterone reading depends heavily on what they did the previous two weeks. A low number after a hard rotation may be a completely healthy response.

Numbers create pressure. In a culture that prizes physical capability, telling someone their testosterone is "low" is not a neutral act. It creates real pressure to seek treatment, whether or not there is a medical reason. That is how normal human variation gets turned into a condition.

The guidelines say don't. The Endocrine Society recommends against population screening in civilians. Soldiers are, if anything, a harder case, not an easier one.

Women are in the military too. Testosterone matters for women's health as well, but circulating levels run about ten times lower. That requires entirely separate reference ranges and a different interpretation framework. How the policy handles this is not yet clear.

The distinction that matters most

There are two completely different things people mean when they say "testosterone treatment."

Treating a deficiency means a man has genuinely low testosterone, has symptoms, has been properly diagnosed with repeat morning tests, and receives treatment to restore normal function. This is established, evidence-based medicine. Nobody is arguing against it.

Pursuing optimization means giving testosterone to someone whose body is working fine, hoping to push past their natural range and get better performance. The evidence that this works is weak, and the risk-versus-benefit math is genuinely unsettled.

And when low testosterone is being caused by poor sleep, not eating enough, training too hard, or chronic stress, then fixing those things is not the boring alternative to treatment. It is the treatment. It addresses the actual cause instead of overriding the signal.

The bottom line

Testosterone is a useful health marker that probably deserves more attention in regular medical care than it gets. That part of the argument holds up.

But a blood test cannot tell you whether someone can fight. It cannot reliably tell you who needs treatment, especially in a population whose levels swing wildly based on last week's schedule. And it cannot tell you whether adding more hormone will make anyone better at their job, because the one study that tested exactly that found it did not.

What the evidence supports is targeted testing of people who actually have symptoms, followed by a real workup from someone who knows what they are looking at.

What it does not support is the assumption underneath the slogan: that more testosterone equals more capability. The body is not a video game character. You cannot just drag the strength slider to the right.

This article is general education, not medical advice, and it describes a current government policy whose details may change after publication. On the science: treating genuinely low testosterone in men with real symptoms and a proper diagnosis is established medicine and not in dispute. What's unsettled is population-wide screening of people who feel fine, and "optimization" past a normal range, where the evidence for real-world benefit is weak and the risks are still being mapped. If you're wondering about your own levels, the meaningful path is a symptom-based conversation with a doctor and repeat morning blood tests, not a single number, and it's worth ruling out sleep, weight, and stress as the actual driver first. For the fuller picture on diagnosis and treatment, see our other pieces on low testosterone.