The label said clinically proven. The evidence said otherwise.
“Ninety percent of testosterone boosters claimed to raise T. When researchers checked the evidence behind every ingredient, three out of four had nothing.”
The ad hits during a scroll. A man, roughly your age, looking slightly more put-together than you feel lately. The caption says natural testosterone support — clinically proven, 30-day supply. The reviews are solid. The price says serious product.
And somewhere in the back of your mind, a question that's been sitting there longer than you'd admit: maybe the gains aren't coming like they used to because something's dropping.
A urology research team at the University of Southern California had the same question — but they brought a different tool. Chase Clemesha and colleagues didn't order one bottle. They pulled 50 testosterone booster supplements off Google's first pages and cataloged every ingredient on every label.
One hundred and nine unique components across 50 products. Then they did what no consumer has the time or training to do — they searched the medical literature for every single component, asking one question per ingredient: does any published evidence show this affects testosterone?
Their results, published in the World Journal of Men's Health in 2020, read like an audit of an industry that never expected to be audited.
Researchers checked 50 testosterone boosters ingredient by ingredient against the medical literature. The gap between what the labels promise and what the science shows is not close.
- For 61.5% of ingredients in testosterone boosters, nobody has ever published a single study on whether they affect testosterone — the products are selling ingredients with no evidence at all.
- Thirteen products exceeded FDA upper safety limits for zinc, vitamin B3, or magnesium — doses the government specifically flags as potentially harmful.
- Even among the 25% of ingredients with some evidence of raising testosterone, more than half also had conflicting data showing no change or a decrease.
- Three ingredients — ashwagandha, fenugreek, and D-aspartic acid — survived the audit with evidence on both sides. Whether that evidence holds up under individual scrutiny is a different question.
What the Receipts Showed
Ninety percent of the 50 products claimed to boost testosterone. Forty-five out of fifty bottles, each carrying the same promise in some variation — boost T, increase free testosterone, support healthy testosterone levels.
When Clemesha's team checked those claims against the medical literature, ingredient by ingredient, the receipts didn't match.
Only 24.8% of the 109 individual components had any published data showing a testosterone increase. Not strong data. Not definitive data. Any data at all. A single study suggesting the ingredient might move the needle counted.
And the rest? 61.5% of components had zero published data on their effect on testosterone. Nobody had ever checked. Not in animals, not in cells, not in humans. The label says "testosterone support" next to an ingredient that has never been tested for its effect on testosterone.
Only 5.5% of all ingredients had been examined in more than two studies. The evidence base, where it existed at all, was paper-thin.
This wasn't a margin of error. The distance between what the labels claimed and what the literature showed was a canyon — 90% of products claiming, 25% of ingredients with any data, 62% with nothing.
Four years later, a completely independent European urology team reviewed 52 published studies across 27 testosterone booster ingredients [1]. Different methodology. Different continent. Their conclusion landed in the same place: most testosterone booster ingredients fail to increase total testosterone. Two research teams, different methods, different continents, same answer.
The Ingredient Working Against You
If the gap between claims and evidence were the whole story, this would be a tale of wasted money and misplaced trust. It gets worse.
Clemesha's team found that 10.1% of the 109 components had published data showing a decrease in testosterone with supplementation. Not zero effect — a move in the opposite direction from what the label promised.
The researchers' own language is careful. They wrote that it is "even more concerning that some of these supplements may in fact decrease serum T." The word may matters. These are published findings from individual studies, not a definitive clinical verdict on each ingredient. But the direction is real.
The person who bought the product to raise their testosterone could be putting something in their body that published research suggests might lower it. One in ten ingredients, pointing the wrong way on the very measure the bottle was bought to improve.
“One in ten ingredients in testosterone boosters had published data showing a decrease in the very hormone the product promises to raise.”
Thirteen Hundred Percent of Your Daily B12
The testosterone question is the headline. But the research team found something else sitting inside these products that raises a different kind of concern entirely.
The median testosterone booster in the analysis contained 1,291% of the recommended daily allowance for vitamin B12. Thirteen times the daily recommendation, packed into a product that isn't sold as a vitamin supplement.
Vitamin B6 came in at 807.6% of the recommended dose. Zinc at 272%. Thirteen of the fifty products exceeded the FDA's upper tolerable intake limits for at least one ingredient — zinc, vitamin B3, or magnesium. These are the doses the FDA specifically flags as potentially harmful.
The researchers noted that even relatively low levels of zinc over-supplementation can interfere with copper and iron use in the body and affect cholesterol. Higher doses can lead to anemia and weakened immune function. Their conclusion: these vitamin and herbal substances may not be as safe as the public perceives them to be.
In December 2023, the FDA discovered that a testosterone booster sold on Amazon contained hidden tadalafil — the active ingredient in Cialis [2]. A consumer buying a "natural" testosterone supplement was unknowingly taking a prescription erectile dysfunction drug that can cause dangerous blood pressure drops.
The FDA states it cannot test all dietary supplements for hidden ingredients. The caught cases are the ones we know about.
Billions In, Evidence Out
The global testosterone booster supplement market was valued at $3.74 billion in 2024, growing at 10.4% annually, with nearly half of all sales happening online [3]. The same platforms where the ad shows up. The same shelves — virtual or physical — where the purchase decision happens.
The market is getting bigger. The evidence base is not. More money, more marketing, more claims — same 24.8% foundation of published data underneath.
But the data resists a blanket verdict. That 24.8% is not zero. Clemesha's own analysis showed a handful of ingredients — ashwagandha, fenugreek, D-aspartic acid — appearing on the evidence side of the ledger. They had published studies. They had data showing a testosterone response, at least in some contexts.
When separate research teams examined these ingredients in isolation — ashwagandha across 13 studies [4], fenugreek across four clinical trials [5], D-aspartic acid across 27 studies including four in humans [6] — each told a story more complicated than any label captures. The category failed the audit. A few ingredients survived it.
“Two research teams on different continents, using different methods, checked the same supplement category four years apart. Both reached the same conclusion.”
What the Shelf Looks Like Now
Clemesha's analysis carries limitations the researchers themselves named. Only Google was used to find products. Only 50 supplements were evaluated — a snapshot of the market, not the full inventory.
The evidence review treated all published studies equally, without distinguishing animal data from human clinical trials. Product formulations change. The specific bottles analyzed in 2020 may not match what's on the shelf today.
None of those caveats close the canyon. Even granting every limitation, the gap between what labels claim and what the literature shows remains the product of a systematic check anyone could replicate. The caveats affect precision. They don't change direction.
The next time the ad shows up during a scroll — natural testosterone support, clinically proven, 30-day supply — the phrase "clinically proven" carries different weight. Not zero weight. But different weight. The label lost something it can't get back, and the person reading it gained something the label never meant to give them.
Three ingredients kept their names on the evidence side of the ledger. Whether they hold up under dedicated scrutiny — ashwagandha tested in a Bayesian meta-analysis, fenugreek measured across randomized trials, D-aspartic acid tracked in resistance-trained men who actually lift — is a different question from whether the category holds up. And the answers are not what the marketing would predict.
The next time a testosterone booster ad promises "clinically proven" results, that phrase now carries a price tag the label can't see. Clemesha's data showed that the clinical evidence behind most of those claims is either missing entirely or pointing in the wrong direction.
The study didn't test whether supplements work — it tested whether the labels tell the truth about what's inside them. That's a question any buyer can check for themselves: pick up the bottle, read the ingredient list, and search for each component in the published research. The gap between what you find and what the bottle promised is the same gap the researchers measured.
What changes after this data isn't which product to buy. It's what the phrase "clinically proven" is worth on a label that has never been required to prove it.
What other research found
What this means for you
The first thing to check is what's actually in the bottle. Clemesha's team found that the average testosterone booster contained 8.3 different ingredients — and for most of those, nobody has ever published a study on whether they affect testosterone.
The safety question is worth checking too. Thirteen of the fifty products exceeded the FDA's upper tolerable limits for at least one vitamin or mineral. The ingredient label tells you the doses — the FDA publishes the limits. Whether your product crosses that line takes about two minutes to verify.
The evidence gap Clemesha documented isn't about finding the right product — it's about the category itself. Most products made the same promise, and for the majority of their ingredients, nobody has ever published a study on whether they affect testosterone at all.
Before spending money, the data suggests one question: can you find a published human study for the specific ingredients on the label? If the ingredient doesn't show up in PubMed with testosterone data, you know more about the evidence base than the label told you.
The study's numbers reframe the conversation. What looked like a reasonable recommendation — the product had good reviews, the ingredients sounded scientific — now sits on a foundation where even the ingredients with published data had conflicting results more than half the time.
The most useful thing to share isn't a verdict on every product. It's the gap itself: most testosterone booster claims have never been checked against published research. That single fact changes what "clinically proven" means on the label.
Before you change anything
This study analyzed products, not people. The 50 testosterone boosters came from Google's first search results — meaning they represent what a US-based buyer would find when shopping online, not the full global market.
If you buy supplements through a different search engine, a specialized retailer, or in a country with different regulations, your product landscape may look different. The evidence gap applies to what PubMed says about each ingredient, which is global — but the specific products analyzed were a snapshot of the US online market in 2020.
The study didn't test these products in humans. It checked what the labels claim against what published research shows. Whether a specific product raises testosterone in a specific person requires a different kind of study entirely.
The PubMed search treated all study types equally. An ingredient marked as having "evidence of testosterone increase" might have that evidence from a single rat study — not from a human clinical trial. The 24.8% figure doesn't distinguish between strong and weak evidence.
Product formulations change frequently. The specific bottles analyzed in 2020 may not match what's on the shelf today. Ingredients get added, removed, or re-dosed between production runs.
The study used a single search term ("Testosterone Booster") on a single search engine. Different search terms or platforms might surface a different product mix — though the underlying PubMed evidence for each ingredient doesn't change with how you find the product.
The category-level verdict is strong. A descriptive analysis counting ingredients against published data doesn't require complex statistics to interpret — the counting itself is the evidence. And a completely independent European team reached the same conclusion four years later using different methods.
The ingredient-level picture is more uncertain. Some of the 24.8% with evidence have robust human data (ashwagandha, fenugreek). Others have a single animal study that might not translate to humans at all. The study counts evidence presence, not evidence quality — a distinction that matters for any ingredient-specific decision.
The convergent pattern (two teams, different methods, same answer) makes the category-level finding reliable enough to change how a buyer evaluates label claims. The ingredient-specific details warrant the dedicated analysis each satellite provides.
The category failed the evidence check. But three ingredients kept their names on the right side of the ledger — and each one tells a story the label couldn't fit.
Ashwagandha showed up in a Bayesian meta-analysis covering 13 studies. Fenugreek raised testosterone in four clinical trials — but nobody measured whether the number on the blood test translated into anything a lifter would notice. D-aspartic acid, the staple of budget test boosters, revealed a dose curve that runs in the opposite direction from what buyers expect.
The survivors of the audit get their own dedicated scrutiny.
What This Study Found
All findings from this paper, in plain language.
- Nine out of ten testosterone boosters claimed to raise testosterone, but only one in four ingredients had any published research backing that claim.
- For nearly two-thirds of ingredients, nobody had ever published a single study on whether they affect testosterone at all.
- About one in ten ingredients had published data showing a testosterone decrease — the opposite direction from what the product promised.
- Even among ingredients with some positive evidence, more than half also had conflicting studies showing no change or a decrease.
- D-aspartic acid — one of the top ten most common ingredients — had evidence pointing in all three directions: increase, decrease, and no change.
- Ashwagandha, found in about 18% of products, had evidence of both a testosterone increase and no change depending on the study.
- Fenugreek extract — the second most common ingredient — was one of the few with only positive testosterone data and no conflicting results.
- The typical testosterone booster contained over 1,200% of the recommended daily B12 dose and over 800% for vitamin B6.
- Thirteen products exceeded government safety limits for at least one ingredient — zinc, vitamin B3, or magnesium.
- Zinc was the single most common ingredient, appearing in nearly two-thirds of products, at almost three times the recommended daily dose.
- The products made 16 different types of claims — from boosting energy to burning fat to improving sleep — most with little to no supporting evidence.
- Only 5.5% of ingredients had been studied more than twice for their effect on testosterone — the evidence base was thin even where it existed.
- Published case reports linked testosterone booster use to serious adverse events including blood clots and liver damage, though these were individual cases.