Raising testosterone without medication
Weight, sleep, alcohol, training and the supplement aisle — graded honestly, including the interventions that do nothing at all.
The interventions that reliably raise testosterone are the ones nobody sells.
Weight loss. Sleep. Eating enough. Reviewing your medications. Treating the disease that’s suppressing your axis. None of them come in a bottle, none carry a margin, and all of them outperform every product marketed as a testosterone booster.
That is not a rhetorical flourish. It is what the evidence shows, and this page walks through it intervention by intervention with a grade attached to each.
Weight loss
🟢 Established. The single most effective non-pharmacological intervention.
A meta-analysis of 24 studies found weight loss through a low-calorie diet raised total testosterone by an average of 2.87 nmol/L (95% CI 1.68–4.07) — roughly 83 ng/dL. Weight loss after bariatric surgery raised it by 8.73 nmol/L (95% CI 6.51–10.95) — roughly 251 ng/dL. The amount of weight lost was the strongest predictor of how much testosterone rose.
A 2025 study of 69 hypogonadal men undergoing bariatric surgery found mean testosterone rising from 208 to 371 ng/dL, with 45% reaching eugonadal status.
Two honest caveats. Fifty-five percent of those men did not normalize despite major surgical weight loss — some men have obesity and an independent cause. And there is no verified threshold of weight loss that predicts normalization for any individual man.
The mechanism is mostly aromatase. Adipose tissue converts testosterone to estradiol, and estradiol feeds back to suppress the pituitary. Less fat, less conversion, less suppression. Insulin resistance also suppresses SHBG, which depresses total testosterone independently.
For the pharmacological route to the same endpoint, see Obesity, GLP-1 Medications and Testosterone.
Sleep
🟡 Promising, and badly under-asked-about.
In ten healthy young men, one week of five hours’ sleep per night lowered daytime testosterone by 10 to 15% — a change the authors compared to ten to fifteen years of normal aging.
Ten men, young, one week. It is a small study and the only clean experimental one we could verify. There is no established dose-response curve across sleep durations — nobody has properly compared six versus seven versus eight hours.
But testosterone secretion is sleep-dependent, chronic sleep restriction is extremely common, and it independently produces fatigue, low mood, poor concentration and reduced libido. A man sleeping five hours a night has a plausible explanation for both his symptoms and his number.
Sleep apnea — treat it, but not for this
🟢 Established that it needs treating · ⚫ Unsupported that treating it raises testosterone.
Obstructive sleep apnea travels with low testosterone. It also produces the entire symptom cluster men attribute to low testosterone.
But a meta-analysis of seven studies across 232 men found CPAP had no significant effect on testosterone — total testosterone SMD −0.14 (95% CI −0.63 to 0.34, p=0.558), and restricted to the randomized trials, −0.05 (p=0.757). The authors concluded CPAP “has no influence on testosterone levels in men with OSA.”
Much of the association probably runs through obesity, which causes both.
Get the apnea treated. It is dangerous, it causes your symptoms, and it may be the actual diagnosis. Just don’t expect your testosterone to move.
Resistance training
🟠 Limited for baseline testosterone — and the mechanism most people believe in is wrong.
Here are two things that are both true and sound contradictory.
Heavy resistance exercise produces an acute testosterone spike. It peaks roughly 15 to 30 minutes after training and returns to baseline within about an hour. This is real and reproducible.
That spike does not build muscle. This is the finding that matters, and almost nobody knows it. West and Phillips ran a within-subject design comparing training performed in a high-hormone versus a low-hormone environment. Muscle cross-sectional area increased 12% in the low-hormone condition versus 10% in the high-hormone condition — no significant difference (p=0.25). Type I fibres: +9% versus +11%. Type II: +21% versus +24%. Isometric strength: +20% versus +19%. One-rep max: +23% versus +25%. None significantly different. Their conclusion was that elevations in ostensibly anabolic hormones with resistance exercise enhance neither hypertrophy nor strength.
The muscle you build from lifting comes from mechanical tension and local signalling, not from a post-workout hormone surge. Which means chasing the spike — training in a particular way to maximize it, timing supplements around it — is chasing something that doesn’t do what it’s advertised to do.
Does chronic training raise baseline testosterone in already-healthy men? Probably little to nothing. We located the relevant meta-analysis but could not retrieve its effect sizes, so we are not quoting numbers. Train because training is good for you. Not because it will move your testosterone.
Aerobic exercise
🟡 Promising in obese men, mediated by weight loss.
In obese men and men with type 2 diabetes, aerobic training raises testosterone with a pooled effect size of 0.565 (95% CI 0.307–0.822, p<0.001) across three studies. That effect is substantially mediated by fat loss and improved insulin sensitivity rather than by exercise itself.
At the other extreme, sustained high-volume endurance training is associated with genuinely low testosterone — a published phenomenon sometimes called the exercise-hypogonadal male condition, described in ultra-endurance athletes. It is real. We could not verify a consistent magnitude and are not quoting one.
More is not better. The curve has two ends.
Eating enough
🟠 Limited, and the male literature is thinner than the fitness world assumes.
Severe energy deficit suppresses the reproductive axis in men as it does in women. Male athletes and soldiers under sustained restriction show suppressed testosterone, including a reported case of castrate-level testosterone after prolonged military training.
A 2024 review in Endocrine Reviews is candid that results across studies of male athletes conflict, and that the male-specific literature is limited compared with the equivalent research in women. We could not find a reliable recovery timeline once intake is restored.
The practical version: the man who responds to fatigue and falling performance by eating less and training more is applying the intervention most likely to make it worse.
Dietary fat
🟠 Limited — and this is a case where a popular claim rests on a much weaker foundation than its confidence suggests.
A 2021 systematic review and meta-analysis found low-fat diets associated with lower testosterone: total testosterone standardized mean difference −0.38 (95% CI −0.75 to −0.01, p=0.04), free testosterone −0.37 (p=0.005), DHT −0.30 (p=0.03).
Now the caveats, which are substantial enough that they change the conclusion.
The entire evidence base is 206 men across six crossover studies published between 1979 and 2005. Five of the six were not randomized. The “low fat” arms averaged 19.5% of calories from fat, with one as low as 6.8% — far below what any clinician means by a low-fat diet, which is typically 25 to 30%. And a larger, more modern comparison found a difference of around half a nanogram per decilitre, which is nothing.
So: extremely low fat intake may lower testosterone modestly, on dated and mostly non-randomized evidence, at fat intakes below what almost anyone actually eats. This is not a reason to add butter to your coffee.
Protein
🟢 Established that normal high-protein diets are fine.
A 2022 meta-analysis found very high protein intake — above 3.4 g/kg/day — associated with a total testosterone decrease of around 5.23 nmol/L. Below that threshold, no consistent decrease was found.
For context: average population intake is around 1.3 g/kg/day, and even aggressive bodybuilding diets typically run 1.8 to 3 g/kg. The threshold sits above what essentially anyone eats.
If you have seen the claim that high protein lowers testosterone, this is where it comes from, and it does not apply to you unless you are eating well over three grams per kilogram daily.
Vitamin D
🟢 Established in deficiency · ⚫ Unsupported in replete men.
This distinction is elided in nearly all supplement marketing, and it is the whole story.
A randomized controlled trial gave 98 healthy men 20,000 IU of vitamin D weekly for twelve weeks. No significant effect on total testosterone — median change 0.5 nmol/L, p=0.497. Vitamin D also worsened insulin sensitivity in that trial (p=0.034).
Repletion in men who are genuinely deficient may modestly help. We could not verify the specific figures from the positive trials and are not quoting them. Supplementing a man who is already replete does not raise testosterone.
Check the level. Treat deficiency. Don’t take it as a testosterone intervention.
Zinc
🟢 Established in deficiency · ⚫ Unsupported in replete men.
Same pattern, and the underlying study is worth looking at directly, because it is a good lesson in how a supplement claim gets built. Prasad and colleagues (Nutrition, 1996) restricted zinc in four young men for twenty weeks and reported testosterone falling from 39.9 to 10.6 nmol/L, then supplemented nine marginally deficient older men for six months and reported a rise from 8.3 to 16.0 nmol/L. Both arms were uncontrolled, and the restriction arm’s starting value of 39.9 nmol/L is roughly 1,150 ng/dL — far above the normal male range, which makes the percentage fall dramatic and hard to interpret. We are naming those numbers rather than repeating them as evidence.
Zinc deficiency is uncommon in men eating a varied diet in developed countries. If you have it, correct it. If you don’t, zinc is not a testosterone intervention.
Alcohol
🟡 Promising for chronic heavy use · 🟠 Limited on dose-response.
Chronic heavy drinking suppresses the axis through direct testicular toxicity, hepatic effects on SHBG and hormone metabolism, and central suppression.
Where “heavy” begins is not established, and we could not verify a dose-response curve. One counterintuitive published finding: acute low-dose alcohol may transiently raise testosterone — a reminder that acute and chronic effects differ and that single observations mislead.
Smoking
Worth knowing, and it goes the direction you don’t expect.
Multiple observational studies report that smokers have higher testosterone than non-smokers, with proposed mechanisms involving LH pulsatility and SHBG. We could not verify the magnitude or confirm the mechanism from primary sources.
This is not an argument for smoking. It is an argument for being suspicious of testosterone as a health metric. A behaviour that will take years off your life is associated with a higher number on the test that men are being told measures their vitality.
Stress
🟠 Limited.
Glucocorticoids suppress the HPG axis, and severe physiological stress clearly lowers testosterone. Everyday psychological stress is much less well quantified, and we could not find a dose-response relationship worth publishing.
Plausible contributor. Not a documented cause of clinical hypogonadism at ordinary levels.
Testosterone boosters: what actually works?
The short version: with the partial exception of correcting a documented zinc or vitamin D deficiency, no supplement marketed as a testosterone booster has been shown to raise testosterone meaningfully in men who are not deficient in a specific nutrient.
Below, individually.
| Supplement | Best human evidence | Grade |
|---|---|---|
| Ashwagandha | Small RCTs, around 50–60 men over 8 weeks, show modest testosterone increases and improved sexual well-being scores. Small effects, mixed trial quality, industry funding common. | 🟠 Limited |
| Tongkat ali | Small RCTs show modest free testosterone increases mainly in men with low-normal baseline. Long-term hepatic safety data are sparse and there are case-report-level hepatotoxicity concerns. | 🟠 Limited + safety flag |
| Fenugreek | Several small trials on proprietary extracts, mostly sponsor-funded, showing modest increases. Weak independent replication. | 🟠 Limited |
| Zinc | Clear benefit correcting deficiency. No consistent effect in replete men. | 🟢 Established in deficiency / ⚫ Unsupported otherwise |
| Vitamin D | A one-year randomized trial in 98 men with 25(OH)D below 75 nmol/L (mean 53.3) and normal testosterone found no effect — median change 0.5 nmol/L, p=0.497. An earlier trial in deficient, overweight men on a weight-loss programme reported a within-group rise, but no between-group treatment effect. | 🟠 Limited even in deficiency / ⚫ Unsupported otherwise |
| Magnesium | Small, inconsistent studies, mostly in exercising men. | 🟠 Limited |
| D-aspartic acid | The best trial — resistance-trained men, 3 g and 6 g daily for 12 weeks — found no significant effect. The widely-cited positive study was in a different, untrained population. | ⚫ Unsupported |
| Boron | Very small trials, fewer than 20 men, suggesting modest free testosterone rise and SHBG reduction. Badly underpowered. | 🟠 Limited |
| Tribulus terrestris | A 2025 systematic review found the evidence for raising testosterone weak and inconsistent. Some erectile function benefit appears independent of any testosterone change. | ⚫ Unsupported for testosterone |
| Maca | Libido improvements reported without consistent testosterone change — so whatever it does, it isn’t hormonal. | ⚫ Unsupported for testosterone / 🟠 Limited for libido |
| Fadogia agrestis | No published human trials at all. Every claim traces to rodent work — which also raised concerns about testicular toxicity at higher doses. | ⚫ Unsupported + safety flag |
| Shilajit | Two small trials report modest increases with purified product. Heavy metal contamination — lead, arsenic — is documented in poor-quality product. | 🟠 Limited + purity flag |
| Proprietary “test booster” blends | Typically combine several of the above at doses below those studied, with no trial of the blend itself. | ⚫ Unsupported |
Three that deserve a specific safety warning
Fadogia agrestis. There are no human trials. None. The entire evidence base is rodent studies, and among those rodent studies is one raising concerns about testicular toxicity at higher doses. You would be taking a compound with zero human safety data and an animal signal pointing at the organ you are trying to help.
Tongkat ali. Case-report-level hepatotoxicity concerns exist, and product purity is variable.
Shilajit. Documented heavy metal contamination in unpurified product.
Consumer coverage of these three almost universally discusses efficacy and skips safety. That asymmetry is not accidental.
What this section can’t do
Lifestyle change works well for men whose testosterone is low because of something modifiable. It does nothing for a man with Klinefelter syndrome, a pituitary adenoma, or a testis lost to torsion.
If your testosterone is low because of structural or genetic disease, no amount of sleep, protein or deadlifting will change it, and time spent pursuing that is time not spent on treatment that works. That is why Why Is Testosterone Low? comes before this page in the reasoning, even though it comes after it in most men’s search history.
What remains uncertain
How much weight loss any individual man needs. The dose-response is established at population level and not at individual level.
Whether any specific training mode raises baseline testosterone. A meta-analysis of 11 randomized trials in 421 previously inactive men found exercise training had no effect on resting testosterone — standardized mean difference 0.00, 95% CI −0.20 to 0.20 — with no difference by mode, age or body-mass status (Potter and colleagues, J Strength Cond Res, 2021). What is missing is a resistance-training-specific meta-analysis; Potter’s mode subgroup is the closest thing to one.
The dose-response for sleep. Only one experimental restriction paradigm exists, in ten young men.
Whether very low fat intake matters at real-world intakes. The evidence base is 206 men, mostly non-randomized, from studies up to 45 years old, at fat intakes almost nobody eats.
The magnitude of the smoking association and why it exists.
Alcohol dose-response. Not established.
Recovery timelines from energy-deficit suppression in men. Not characterized.
Whether any supplement helps a specific subgroup. Most trials are small, short, and often funded by the manufacturer.
Questions patients ask
Testosterone boosters work.
With the exception of correcting a real nutrient deficiency, no.
What the evidence showsNo supplement marketed as a testosterone booster has been shown to raise testosterone meaningfully in men who are not deficient in zinc or vitamin D. The trials that exist are small, short, frequently sponsor-funded, and often show effects that don't replicate.
What remains uncertainWhether some help specific subgroups. Nobody has shown it.
Bottom lineThe category is marketing, not medicine.
Strong
Losing weight raises testosterone.
Yes, more than any other non-medical intervention.
What the evidence showsMeta-analysis of 24 studies: low-calorie dieting raised testosterone by 2.87 nmol/L (~83 ng/dL); bariatric surgery by 8.73 nmol/L (~251 ng/dL). Degree of weight loss was the strongest predictor.
What remains uncertainIn a 2025 study, 45% of hypogonadal men reached eugonadal status after bariatric surgery — meaning 55% didn't.
Bottom lineThe highest-yield thing you can do. Not a guarantee.
Strong
Lifting weights raises your testosterone.
Acutely yes, for about an hour. Chronically, probably not much — and the acute spike doesn't do what people think.
What the evidence showsThe post-exercise rise peaks 15–30 minutes after training and resolves within about an hour. A within-subject study comparing training in high- versus low-hormone conditions found muscle cross-sectional area up 12% versus 10% — no significant difference (p=0.25) — with strength and fibre-type results similarly indistinguishable.
What remains uncertainWhether chronic training changes baseline testosterone in healthy men. We could not verify effect sizes.
Bottom lineLift because lifting works. The hormone spike isn't why it works.
Strong on the hypertrophy point
Sleep affects testosterone.
Substantially.
What the evidence showsOne week of five hours' sleep per night lowered daytime testosterone by 10 to 15% in ten healthy young men — comparable to ten to fifteen years of aging.
What remains uncertainSmall study, young men. No dose-response across sleep durations.
Bottom lineOne of the most plausible reversible causes and among the least often asked about.
Moderate
Zinc raises testosterone.
Only if you're deficient.
What the evidence showsExperimental zinc restriction lowers testosterone and repletion in deficient men raises it. No consistent effect in men who are already replete.
What remains uncertainWe could not retrieve the numbers from the foundational study and cite it qualitatively.
Bottom lineCorrect a deficiency. Don't take it as a booster.
Established in deficiency, Unsupported otherwise
Vitamin D raises testosterone.
Same answer as zinc, with a clean negative trial to back it.
What the evidence showsA randomized trial gave 98 healthy men 20,000 IU weekly for twelve weeks and found no significant effect on total testosterone (p=0.497). It also worsened insulin sensitivity.
What remains uncertainThe magnitude of benefit from repleting genuinely deficient men.
Bottom lineCheck your level. Treat deficiency. Otherwise it does nothing here.
Established in deficiency, Unsupported otherwise
Ashwagandha raises testosterone.
Possibly, modestly, on weak trials.
What the evidence showsSmall randomized trials of around 50–60 men over eight weeks report modest increases and improved sexual well-being scores. Effect sizes are small, quality is mixed, and industry funding is common.
What remains uncertainWhether the effect replicates outside sponsor-funded trials, and whether it's clinically meaningful.
Bottom lineThe best of a weak field. Not close to a treatment.
Limited
Tongkat ali works.
Modest signal in men with low-normal levels — and a safety question worth taking seriously.
What the evidence showsSmall trials show modest free testosterone increases mainly in men with low-normal baseline. Long-term hepatic safety data are sparse, with case-report-level hepatotoxicity concerns, and product purity varies.
What remains uncertainEfficacy in men with normal levels, and the real hepatic risk.
Bottom lineWeak efficacy signal, non-trivial safety question, unregulated supply.
Limited + safety flag
Fadogia agrestis is a natural testosterone booster.
There are no human trials, and the animal data include a testicular toxicity signal.
What the evidence showsEvery claim traces to rodent studies. Among those rodent studies is one raising concerns about testicular toxicity at higher doses. No published human trial exists.
What remains uncertainEverything, including whether it is safe.
Bottom lineZero human safety data and an animal signal pointing at the testes. This one we would actively avoid.
Unsupported + safety flag
D-aspartic acid raises testosterone.
The best trial says no.
What the evidence showsA trial in resistance-trained men using 3 g and 6 g daily for twelve weeks found no significant effect on total or free testosterone. The frequently cited positive study was in a different, untrained population.
What remains uncertainWhether untrained or deficient men respond differently. Not demonstrated.
Bottom lineWell studied, and it doesn't work.
Unsupported
Tribulus raises testosterone.
No, though it may do something for erectile function by another route.
What the evidence showsA 2025 systematic review found the evidence for raising testosterone weak and inconsistent. Some erectile function benefit appears independent of testosterone change.
What remains uncertainWhat mechanism produces the erectile signal, if it is real.
Bottom lineNot a testosterone intervention.
Unsupported for testosterone
A low-fat diet tanks your testosterone.
Very low fat intake may lower it modestly, on evidence much weaker than the claim's confidence implies.
What the evidence showsA 2021 meta-analysis found a standardized mean difference of −0.38 (95% CI −0.75 to −0.01, p=0.04). The basis is 206 men across six crossover studies from 1979 to 2005, five of which were not randomized, with "low fat" arms averaging 19.5% of calories — far below normal dietary advice. A larger modern comparison found a difference of about half a nanogram per decilitre.
What remains uncertainWhether any of this applies at real-world fat intakes.
Bottom lineDon't eat 7% fat. Beyond that, this isn't a lever.
Limited
High protein lowers testosterone.
Only at intakes essentially nobody eats.
What the evidence showsA 2022 meta-analysis found a decrease of around 5.23 nmol/L only above 3.4 g/kg/day. Below that — which includes every normal high-protein diet — no consistent decrease. Average intake is around 1.3 g/kg/day.
What remains uncertainWhy the effect appears at the extreme.
Bottom lineYour protein intake is not the problem.
Moderate
Does alcohol lower testosterone?
Chronic heavy use does. The threshold isn't established.
What the evidence showsChronic heavy drinking suppresses the axis through testicular toxicity, hepatic effects and central suppression. Acute low-dose alcohol has been reported to transiently raise it.
What remains uncertainThe dose-response. We could not verify one and won't invent a threshold.
Bottom lineHeavy drinking is a plausible cause. Where heavy starts, nobody has established.
Limited
Quitting smoking will raise my testosterone.
Probably the opposite, which tells you something about testosterone as a health metric.
What the evidence showsMultiple observational studies report smokers having higher testosterone than non-smokers, with proposed mechanisms involving LH pulsatility and SHBG. We could not verify magnitude or mechanism from primary sources.
What remains uncertainWhy, and by how much.
Bottom lineQuit anyway. A number that goes up when you smoke is not a measure of health.
Limited
Treating my sleep apnea will fix my testosterone.
It probably won't — and you should still treat it.
What the evidence showsA meta-analysis of seven studies in 232 men found CPAP had no significant effect on testosterone, including in the randomized trials alone. Much of the OSA–testosterone association likely runs through obesity.
What remains uncertainHow much of the association is independent of weight.
Bottom lineTreat it because it's dangerous and causes your symptoms. Not for your testosterone.
Moderate
Overtraining lowers testosterone.
Sustained high-volume training with inadequate fuelling does, though the male literature is thinner than the fitness world assumes.
What the evidence showsLow testosterone in high-volume endurance athletes is a published phenomenon. A 2024 review in *Endocrine Reviews* notes conflicting results across studies and a limited male-specific literature.
What remains uncertainThreshold, magnitude, and recovery timeline — none established.
Bottom lineReal, and less well characterized than confident internet advice suggests.
Limited
Under-eating lowers testosterone.
Severe energy deficit does.
What the evidence showsMale athletes and soldiers under sustained restriction show suppressed testosterone, including a case of castrate-level testosterone after prolonged military training. The male literature is limited and conflicting.
What remains uncertainHow much deficit is needed, and how fast recovery happens.
Bottom lineIf you're dieting hard and training hard and feel terrible, eating more is a candidate intervention.
Limited
I can raise my testosterone naturally instead of using TRT.
Sometimes — it depends entirely on why it's low.
What the evidence showsIf the cause is obesity, sleep deprivation, energy deficit, alcohol or a medication, lifestyle change can produce meaningful improvement, with weight loss giving effects comparable in magnitude to replacement. If the cause is Klinefelter syndrome, a pituitary adenoma, or testicular loss, lifestyle change will do nothing.
What remains uncertainHow to predict which men will normalize.
Bottom lineEstablish the cause first. That determines whether this page is relevant to you at all.
Strong
How much can I raise testosterone naturally?
In an obese man, a lot. In a lean healthy man with an organic cause, very little.
What the evidence showsWeight loss produced average rises of 83 ng/dL through dieting and 251 ng/dL after bariatric surgery. Correcting a genuine vitamin D or zinc deficiency helps. Sleep restriction costs 10–15%. Everything else is small or unproven.
What remains uncertainIndividual variability is wide and unpredictable.
Bottom lineThe ceiling is set by what's causing the problem, not by how hard you try.
Moderate
Where to go next
Why your testosterone is low: Why Is Testosterone Low? The pharmacological route to weight loss: Obesity, GLP-1 Medications and Testosterone When something else is the problem: When Testosterone Isn’t the Answer
Ready for testosterone care built on a diagnosis?
The first step is a full endocrine evaluation, not a prescription. We see patients across San Diego County and welcome referrals from other physicians.