Do Testosterone Boosters Work? An Ingredient-by-Ingredient Look at the Evidence

The number that frames the entire category

Before any single ingredient, one study frames the whole question. Clemesha, Thaker and Samplaski analysed the 50 best-selling testosterone-boosting supplements sold online and catalogued every component on their labels. Their findings, published in the World Journal of Men's Health in 2020, are the most useful thing anyone has published about this category.

Industry audit — Clemesha CG, Thaker H, Samplaski MK, World J Mens Health 2020;38(1):115–122 (PMID 31385468)

What was measured	Result
Supplements analysed	50 best-selling online testosterone boosters
Distinct components identified	109
Products claiming to boost testosterone	90%
Components with data supporting a testosterone increase	24.8%
Components with data showing no change	18.3%
Components with data suggesting a negative effect	10.1%
Components with no published testosterone data at all	61.5%

Read those last two rows together. On the labels of the fifty best-selling products in this category, roughly six in ten ingredients had never been studied for the effect they were being sold for, and one in ten had data pointing the wrong way. That is not a claim about any particular brand. It is a description of the shelf.

It also tells you how to read the rest of this article. The right question is never "does a testosterone booster work". It is "does this ingredient, at this dose, in this kind of man, do anything measurable". Below is that question answered nine times.

The verdict matrix: nine ingredients, graded on human data

Every row below is anchored to a named human study or systematic review. Where the evidence is thin, the row says so rather than rounding up.

Ingredient	Best available human evidence	Effect on total testosterone	Effect on libido or other outcomes	Honest verdict
Tongkat Ali (Eurycoma longifolia)	Leisegang K et al., Medicina 2022;58(8):1047 — 9 studies reviewed, 5 RCTs meta-analysed	Pooled SMD 1.352 (95% CI 0.565–2.138, p = 0.001). Heterogeneity very high (I² = 87.3%). Significant in hypogonadal men; not significant in healthy men on subgroup analysis	Talbott SM et al., 2013: at 200 mg/day for 4 weeks, salivary cortisol −16%, salivary testosterone +37%, Tension −11%, Anger −12%, Confusion −15%	The strongest case in the category — and still small trials with a wide confidence interval and an effect concentrated in lower-baseline men
Tribulus terrestris	Vilar Neto JO et al., Nutrients 2025;17(7):1275 — 10 studies, 483 participants	8 of 10 studies reported no significant change in androgen profile. The 2 that did showed modest rises of about 60–70 ng/dL, and only in men who were already hypogonadal	Sexual-function outcomes were assessed in 5 studies at 400–750 mg/day; the review rated this evidence low-level	Authors' own conclusion: no robust evidence was found for increasing testosterone levels
Saw palmetto	CAMUS trial — Barry MJ et al., JAMA 2011;306(12):1344–1351; 369 men aged 45+, dose escalated 320 → 640 → 960 mg/day over 72 weeks	Not a testosterone intervention. No trial evidence that it raises testosterone	Urinary symptom score fell 2.20 points on saw palmetto versus 2.99 on placebo — no advantage. PSA change 0.23 ± 0.83 versus 0.16 ± 1.08 ng/mL on placebo (p = 0.50)	Appears in male formulas for DHT-pathway reasons; the large, well-designed trials that exist were null on their own endpoints
Ashwagandha (Withania somnifera)	Lopresti AL et al., Am J Mens Health 2019;13(2) — crossover RCT, 57 enrolled, overweight men aged 40–70	14.7% greater rise in salivary testosterone than placebo (p = .010); DHEA-S 18% greater (p = .005)	No significant between-group difference in cortisol, estradiol, fatigue, vigour or sexual well-being	A real hormonal signal in one small trial — but the outcomes a man would actually feel did not follow it
Fenugreek	Mansoori A et al., Phytother Res 2020;34(7):1550–1555 (PMID 32048383) — meta-analysis of 4 trials	Pooled analysis reported a significant effect on total serum testosterone	Not the focus of this meta-analysis	Points in a positive direction, but four trials is a thin base to build a purchase on
D-aspartic acid	Melville GW et al., J Int Soc Sports Nutr 2015;12:15 — 24 resistance-trained men, 14 days	6 g/day significantly reduced total testosterone (p = 0.03) and free testosterone (p = 0.005). 3 g/day had no effect on either	Not assessed	At the higher dose the best controlled evidence runs against the marketing claim
Vitamin D	Pilz S et al., Horm Metab Res 2011;43(3):223–225 — 54 completers, 3,332 IU/day for one year	Total testosterone rose from 10.7 ± 3.9 to 13.4 ± 4.7 nmol/L (p < 0.001); no significant change on placebo	Free and bioactive testosterone also rose in the supplemented group	Repletion evidence in overweight men on a weight-reduction programme — not a booster for men who are already replete
Zinc	Prasad AS et al., Nutrition 1996;12(5):344–348 — experimental restriction in 4 young men; supplementation in 9 marginally deficient older men	Induced dietary restriction dropped serum testosterone sharply; supplementation in marginally deficient older men raised it	Not assessed	Classic repletion evidence, and the sample sizes are tiny. Says nothing about zinc-replete men
Magnesium	No adequately powered randomised trial in replete men located	Not established	Not established	Insufficient evidence to call either way. Common on labels; uncommon in the trial literature

How to read a pooled effect size like SMD 1.352

A standardised mean difference is a way of comparing results across trials that measured things slightly differently. By conventional interpretation, 0.2 is small, 0.5 is moderate and 0.8 is large, so 1.352 looks impressive. Two things temper it. The confidence interval runs from 0.565 to 2.138, which is wide, and the reported heterogeneity of I² = 87.3% means the five trials disagreed with each other a great deal. A pooled figure built on five small, heterogeneous trials is a signal worth taking seriously and a number worth holding loosely.

Deficient versus replete: the distinction nobody puts on the label

Almost every argument about testosterone supplements collapses once you separate two completely different situations.

Repletion is fixing a shortfall. If a man is genuinely low in vitamin D or zinc, correcting that can move measured testosterone, because the shortfall was suppressing it. Both the Pilz vitamin D trial and the Prasad zinc work are repletion studies. The vitamin D participants were overweight men in a weight-reduction programme, not a random sample of healthy men, and the zinc work involved four men on a deliberately restricted diet and nine older men who were already marginally deficient.

Supplementation in a replete man is a different question entirely, and the honest answer is that it usually does very little. There is no biological reason why adding more of a nutrient you already have enough of should push a hormone higher, and the trials broadly agree.

The same split shows up in the botanical data, which is why it is the spine of this article. The Medicina meta-analysis found a significant testosterone effect for Tongkat Ali in hypogonadal men and a non-significant one in healthy men. The Nutrients review of Tribulus found the only two positive androgen results in men who were already hypogonadal. Same pattern, two different herbs: the effect appears where there is room for it, and largely disappears where there is not.

The practical consequence is uncomfortable for the category. A man with normal levels, good sleep and a reasonable body composition is the man least likely to notice anything — and he is also the man these products are most heavily marketed to.

What the evidence does not show

This is the section most pages in this category skip, so here it is in plain terms.

• It does not show that raising a number changes how you feel. The clearest example is the ashwagandha crossover trial: salivary testosterone rose 14.7% more than placebo, and yet fatigue, vigour and sexual well-being showed no significant between-group difference. A hormone assay moving is not the same as a life changing.
• It does not show long-term safety or benefit. Almost every botanical trial cited here ran between 4 and 12 weeks in screened participants. Nobody has published what happens at year three.
• It does not show that the finished product works. Ingredients accumulate evidence; multi-ingredient formulas rarely get trialled as sold. Evidence for an ingredient is not evidence for a blend containing that ingredient at an undisclosed amount.
• It does not show equivalence between extracts. The Tongkat Ali trial literature leans heavily on specific standardised extracts. A label that says "Tongkat Ali 200 mg" without naming a standardisation is not automatically the same 200 mg that was studied.
• It does not show anything about treatment. None of these ingredients has been shown to treat hypogonadism or erectile dysfunction, and none should be used as an attempt to. Both are medical conditions with real causes worth diagnosing.
• We could not verify some widely repeated figures. A frequently quoted "2015 FDA study finding no quantifiable benefits" could not be traced to a primary document, so it does not appear above. Neither does the commonly cited "1–2% annual testosterone decline after 30", which we could not tie to a named cohort study. We would rather leave a gap than fill it with a number we cannot stand behind.

So what should a man actually do with this?

Three conclusions follow from the table, and none of them is dramatic.

Fix the shortfalls first. If your vitamin D is low, that is a measurable, correctable thing with a real trial behind it. The same logic applies to sleep and body composition, which move testosterone more reliably than anything in a capsule.

Judge a product on its label, not its ingredient list. The audit numbers at the top of this page exist because the industry sells recognisable names rather than studied doses. If a panel does not tell you how much of each botanical you are getting, you cannot compare it to the trials in the matrix above — and neither can anybody else. Our companion piece on Tongkat Ali versus ashwagandha goes through the same exercise for the two most-stacked botanicals in men's supplements.

Give it a real window, then stop. The botanical trials that reported anything ran 4 to 12 weeks. Judging a capsule after ten days measures the adjustment period, not the intervention. Equally, if twelve honest weeks produce nothing, the evidence says that is your answer. Our pillar review of Tongkat Ali and Tribulus sets out what those ingredients can and cannot reasonably be expected to do.

Frequently asked questions