Testosterone boosters: what the trials really say

There is a scientific review that did exactly what no one does: take all the supplements sold as "testosterone boosters" and look one by one at what the trials say about them. The result can be summed up in one sentence: most do not increase testosterone. Two or three manage to do so in specific situations. Here is which ones, and why nuance matters more than the verdict.

52 studies, 27 ingredients, a single verdict

A systematic review published in 2024 scrutinized twenty years of literature on 27 products sold for testosterone, across four populations: athletes, men with late-onset hypogonadism, infertile men, healthy men. Its conclusion is blunt — most fail. It does, however, name exceptions: HMB and betaine in male athletes, Eurycoma longifolia and ashwagandha in healthy or hypogonadal men.

A second review, in 2021, tackled the problem from the other end: it collected the 32 best-selling online "testosterone" and "erection" supplements, extracted the 37 ingredients, and counted the existing randomized trials for each. One hundred and five trials found — and a rating assigned ingredient by ingredient.

The trials that find nothing

They exist, they are clean, and they are rarely cited — precisely because they find nothing. D-aspartic acid, at 3 g per day for 28 days in already trained men: weight training made them progress, supplementation added nothing, neither to body composition, nor to strength, nor to total or free testosterone. NMDA, tested the same year in the same setup: identical result, no additional effect on anything.

Tribulus terrestris, the most widespread ingredient in the category, was evaluated in 2005 in an unambiguously titled trial: "the aphrodisiac plant Tribulus terrestris does not influence androgen production in young men." Four weeks, 21 men, two different doses — testosterone, androstenedione and LH unchanged, all within normal values.

Tongkat ali in amateur athletes, six weeks at 400 mg per day, did not alter the urinary testosterone/epitestosterone ratio. This is good news for anti-doping control — and bad news for the sales pitch.

The exceptions, and what they are actually worth

Three plants stand out from the two reviews. They are not equal.

Ashwagandha has the most impressive trials. A 2015 randomized trial on 57 young weight-training men, 300 mg twice a day for eight weeks: strength gains clearly superior to placebo, more muscle mass, less fat, and a testosterone increase of 96.2 ng/dL versus 18.0 on placebo. The STAR trial, in 2018, confirmed this regarding strength with another extract:

Fenugreek has a meta-analysis in its corner: four clinical trials combined in 2020 conclude there is a significant effect on total testosterone. But a 2024 trial on 95 men aged 40 to 80 shows how much measurement matters: over twelve weeks, plasma testosterone did rise by 13.0% compared to baseline — without being significantly different from placebo. It is salivary testosterone that clearly increased, by 37.2%. Two ways to measure, two possible conclusions from the same trial.

Eurycoma longifolia, tongkat ali, has the most favorable meta-analysis in the entire category: five randomized trials, a major effect on total testosterone, confirmed in hypogonadal men. It also has a problem that supplement reviews never mention — it is not authorized for sale in Europe, and we explain why in our product pages.

A 2025 review moreover identifies ashwagandha and fenugreek as the two extracts with the best-documented positive effects, using two different mechanisms: ashwagandha reportedly acts by curbing the effect of cortisol and prolactin, while fenugreek acts via the diosgenin it contains, a sex hormone precursor.

The real divide: fixing a deficiency, or exceeding normal levels

This is the line that explains half of the contradictions in the field. A nutrient that is missing and gets replaced makes testosterone go back up. The same nutrient given to someone who is not deficient does nothing.

Zinc demonstrates this in both directions. A benchmark study induced a mild deficiency in young men: their testosterone dropped from 39.9 to 10.6 nmol/L in twenty weeks. In slightly deficient elderly men, supplementation did the reverse — from 8.3 to 16.0 nmol/L. A dietary restriction trial confirms the system's sensitivity: 1.4 mg of zinc per day instead of 10.4 is enough to lower testosterone and semen volume.

Vitamin D tells the exact same story, and it is striking how closely the two trials mirror each other:

This is not a contradiction, it is the same rule seen twice: we only correct what is out of whack. Yet booster marketing targets young, healthy men primarily — that is, the group whose trials show they have nothing to gain.

The case of maca: it works, but not as advertised

Maca deserves a paragraph of its own because it is sold as a hormonal booster and is not one. A 2002 randomized trial over twelve weeks did find an improvement in sexual desire, visible as early as the eighth week — and the article title says the rest: this improvement is unrelated to serum testosterone, which did not budge, nor did estradiol.

A second trial verified this across the entire hormonal panel: LH, FSH, prolactin, progesterone, testosterone, estradiol — nothing changes. Something happens, but not via the pathway put on the label.

What is in the jar, and what can go wrong

A 2020 article focused on the label itself rather than the molecule, by dissecting the claims of a booster. Its finding: scientific vocabulary is hijacked, and the cited research is exaggerated or distorted to make studies say what they do not say.

There are also reports of serious side effects. Acute liver injury in a 27-year-old man after six months of a booster, which improved upon stopping — the article stresses a practical point: the patient had not reported the supplement, which delayed diagnosis. And a renal infarction in a young man, with no usual cause found, attributed to a hypercoagulable state possibly induced by over-the-counter "testosterone" pills.

These are isolated cases, not trials: they do not measure a risk, they signal that it exists. Relative to the number of boxes sold, they remain rare. But they are enough that a "natural" supplement should not be treated as harmless, especially during a long cycle and without talking to your doctor.

What we can take away

Each ingredient cited here has its profile on the site, with its studies listed one by one: tribulus, ashwagandha, fenugreek, maca, zinc, vitamin D. They also say what these studies do not show.

Every figure in this article comes from a study you can open: the underlined mentions lead directly to the summary of the work from which the figure is drawn. These summaries were written from the abstract alone. The trials cited often involve small sample sizes and durations of a few weeks, and several were funded by the manufacturer of the tested extract. Informational content, which does not replace the advice of a professional: low testosterone is diagnosed and treated medically, not down the aisle.